Movatterモバイル変換

Version:	2.7.4
Date:	2025-12-08
Title:	High-Throughput Toxicokinetics
Description:	Pre-made models that can be rapidly tailored to various chemicals and species using chemical-specific in vitro data and physiological information. These tools allow incorporation of chemical toxicokinetics ("TK") and in vitro-in vivo extrapolation ("IVIVE") into bioinformatics, as described by Pearce et al. (2017) (<doi:10.18637/jss.v079.i04>). Chemical-specific in vitro data characterizing toxicokinetics have been obtained from relatively high-throughput experiments. The chemical-independent ("generic") physiologically-based ("PBTK") and empirical (for example, one compartment) "TK" models included here can be parameterized with in vitro data or in silico predictions which are provided for thousands of chemicals, multiple exposure routes, and various species. High throughput toxicokinetics ("HTTK") is the combination of in vitro data and generic models. We establish the expected accuracy of HTTK for chemicals without in vivo data through statistical evaluation of HTTK predictions for chemicals where in vivo data do exist. The models are systems of ordinary differential equations that are developed in MCSim and solved using compiled (C-based) code for speed. A Monte Carlo sampler is included for simulating human biological variability (Ring et al., 2017 <doi:10.1016/j.envint.2017.06.004>) and propagating parameter uncertainty (Wambaugh et al., 2019 <doi:10.1093/toxsci/kfz205>). Empirically calibrated methods are included for predicting tissue:plasma partition coefficients and volume of distribution (Pearce et al., 2017 <doi:10.1007/s10928-017-9548-7>). These functions and data provide a set of tools for using IVIVE to convert concentrations from high-throughput screening experiments (for example, Tox21, ToxCast) to real-world exposures via reverse dosimetry (also known as "RTK") (Wetmore et al., 2015 <doi:10.1093/toxsci/kfv171>).
Depends:	R (≥ 2.10)
Imports:	deSolve, msm, data.table, survey, mvtnorm, truncnorm, stats,graphics, utils, magrittr, purrr, methods, Rdpack (≥ 2.3),ggplot2, dplyr
RdMacros:	Rdpack
Suggests:	knitr, rmarkdown, gplots, scales, EnvStats, MASS,RColorBrewer, stringr, reshape, viridis, gmodels, colorspace,cowplot, ggrepel, forcats, smatr, gridExtra, readxl, ks,testthat, ggpubr, tidyverse
License:	MIT + file LICENSE
LazyData:	true
LazyDataCompression:	xz
Encoding:	UTF-8
VignetteBuilder:	knitr
RoxygenNote:	7.3.3
BugReports:	https://github.com/USEPA/CompTox-ExpoCast-httk/issues
NeedsCompilation:	yes
Packaged:	2025-12-08 16:30:38 UTC; WambaughJohn
Author:	John Wambaugh [aut, cre], Sarah Davidson-Fritz [aut], Robert Pearce [aut], Caroline Ring [aut], Greg Honda [aut], Mark Sfeir [aut], Matt Linakis [aut], Dustin Kapraun [aut], Kimberly Truong [aut], Colin Thomson [aut], Meredith Scherer [aut], Annabel Meade [aut], Celia Schacht [aut], Leonie Lautz [aut], Todor Antonijevic [ctb], Miyuki Breen [ctb], Shannon Bell [ctb], Xiaoqing Chang [ctb], Jimena Davis [ctb], Elaina Kenyon [ctb], Gilberto Padilla Mercado [ctb], Katie Paul Friedman [ctb], Nathan Pollesch [ctb], Noelle Sinski [ctb], Nisha Sipes [ctb], James Sluka [ctb], Caroline Stevens [ctb], Barbara Wetmore [ctb], Lily Whipple [ctb], Woodrow Setzer [ctb]
Maintainer:	John Wambaugh <wambaugh.research@gmail.com>
Repository:	CRAN
Date/Publication:	2025-12-09 06:10:32 UTC

httk: High-Throughput Toxicokinetics

Description

Pre-made models that can be rapidly tailored to various chemicals and species using chemical-specific in vitro data and physiological information. These tools allow incorporation of chemical toxicokinetics ("TK") and in vitro-in vivo extrapolation ("IVIVE") into bioinformatics, as described by Pearce et al. (2017) (doi:10.18637/jss.v079.i04). Chemical-specific in vitro data characterizing toxicokinetics have been obtained from relatively high-throughput experiments. The chemical-independent ("generic") physiologically-based ("PBTK") and empirical (for example, one compartment) "TK" models included here can be parameterized with in vitro data or in silico predictions which are provided for thousands of chemicals, multiple exposure routes, and various species. High throughput toxicokinetics ("HTTK") is the combination of in vitro data and generic models. We establish the expected accuracy of HTTK for chemicals without in vivo data through statistical evaluation of HTTK predictions for chemicals where in vivo data do exist. The models are systems of ordinary differential equations that are developed in MCSim and solved using compiled (C-based) code for speed. A Monte Carlo sampler is included for simulating human biological variability (Ring et al., 2017doi:10.1016/j.envint.2017.06.004) and propagating parameter uncertainty (Wambaugh et al., 2019doi:10.1093/toxsci/kfz205). Empirically calibrated methods are included for predicting tissue:plasma partition coefficients and volume of distribution (Pearce et al., 2017doi:10.1007/s10928-017-9548-7). These functions and data provide a set of tools for using IVIVE to convert concentrations from high-throughput screening experiments (for example, Tox21, ToxCast) to real-world exposures via reverse dosimetry (also known as "RTK") (Wetmore et al., 2015doi:10.1093/toxsci/kfv171).

Author(s)

Maintainer: John Wambaughwambaugh.research@gmail.com (ORCID)

Authors:

• Sarah Davidson-Fritzdavidsonfritz.sarah@epa.gov (ORCID)

• Robert Pearce (ORCID)

• Caroline RingRing.Caroline@epa.gov (ORCID)

• Greg Hondahonda.gregory@epa.gov (ORCID)

• Mark Sfeir

• Matt LinakisMLINAKIS@ramboll.com (ORCID)

• Dustin Kapraunkapraun.dustin@epa.gov (ORCID)

• Kimberly Truongtruong.kimberly@epa.gov (ORCID)

• Colin Thomsonthomson.colin@epa.gov (ORCID)

• Meredith SchererScherer.Meredith@epa.gov (ORCID)

• Annabel Meadeaemeade7@gmail.com (ORCID)

• Celia SchachtSchacht.Celia@epa.gov (ORCID)

• Leonie Lautzleonie.lautz@wur.nl (ORCID)

Other contributors:

• Todor Antonijevictantonijevic@toxstrategies.com (ORCID) [contributor]

• Miyuki Breenbreen.miyuki@epa.gov (ORCID) [contributor]

• Shannon BellShannon.bell@inotivco.com (ORCID) [contributor]

• Xiaoqing ChangXiaoqing.chang@inotivco.com (ORCID) [contributor]

• Jimena Davis [contributor]

• Elaina Kenyonkenyon.elaina@epa.gov (ORCID) [contributor]

• Gilberto Padilla MercadoPadillaMercado.Gilberto@epa.gov (ORCID) [contributor]

• Katie Paul FriedmanPaul-Friedman.Katie@epa.gov (ORCID) [contributor]

• Nathan Polleschpollesch.nathan@epa.gov (ORCID) [contributor]

• Noelle SinskiNoelle.Sinski@icf.com [contributor]

• Nisha Sipessipes.nisha@epa.gov (ORCID) [contributor]

• James Slukajsluka@indiana.edu (ORCID) [contributor]

• Caroline Stevensstevens.caroline@epa.gov (ORCID) [contributor]

• Barbara Wetmorewetmore.barbara@epa.gov (ORCID) [contributor]

• Lily Whipple [contributor]

• Woodrow Setzer (ORCID) [contributor]

See Also

Useful links:

• Report bugs athttps://github.com/USEPA/CompTox-ExpoCast-httk/issues

Test the check digit of a CAS number to confirm validity

Description

Chemical abstracts services registry numbers (CAS-RN) include a final digitas a "checksum" to test for validity (that is, that the number has not beencorrupted).

Usage

CAS.checksum(CAS.string)

Arguments

Details

The check digit (final number) is calculated by working from right to left,starting with the second to last digit of the CAS-RN. We multiply each digitby an increasing digit (1, 2, 3...) and sum as we work from right to left.The check digit should equal the final digit of the sum.

Value

logical (TRUE if final digit of CAS is consistent with other digits)

Author(s)

John Wambaugh

Reference for EPA Physico-Chemical Data

Description

The physico-chemical data in the chem.phys_and_invitro.data table areobtained from EPA's Comptox Chemicals dashboard. This variable indicatesthe date the Dashboard was accessed.

Usage

EPA.ref

Format

An object of classcharacter of length 1.

Author(s)

John Wambaugh

Source

https://comptox.epa.gov/dashboard

A timestamp of table creation

Description

The Tables.RData file is separately created as part of building a newrelease of HTTK. This time stamp indicates the script used to build the fileand when it was run.

Usage

Tables.Rdata.stamp

Format

An object of classcharacter of length 1.

Author(s)

John Wambaugh

Add a table of chemical information for use in making httk predictions.

Description

This function adds chemical-specific information to the tablechem.physical_and_invitro.data. This table is queried by the modelparameterization functions when attempting to parameterize a model, soadding sufficient data to this table allows additional chemicals to bemodeled.

Usage

add_chemtable(  new.table,  data.list,  current.table = NULL,  reference = NULL,  species = NULL,  overwrite = FALSE,  sig.fig = 4,  clint.pvalue.overwrite = TRUE,  allow.na = FALSE,  suppress.messages = FALSE)

Arguments

Value

Author(s)

John Wambaugh

Examples

library(httk)# Number of chemicals distributed with the package:num.chems <- length(get_cheminfo())fake <- data.frame(Compound="Tester",                   CASRN="222-11-1",                   DTXSID="DTX111222",                   MW=200,                   logP=3.5,                   Fup=0.1,                   Clint=0.1,                   Clint.pValue=0.001,stringsAsFactors=FALSE)chem.physical_and_invitro.data <- add_chemtable(  fake,  current.table=chem.physical_and_invitro.data,  data.list=list(    Compound="Compound",    CAS="CASRN",    DTXSID="DTXSID",    MW="MW",    logP="logP",    Funbound.plasma="Fup",    Clint="Clint",    Clint.pValue="Clint.pValue"),  species="Human",  reference="Fake")calc_css(chem.name="Tester")#load_sipes2017()# We should have the ADMet Predicted chemicals from Sipes et al. (2017),# this one is a good test since the logP is nearly 10!#calc_css(chem.cas="26040-51-7")#Let's see how many chemicals we have now with the Sipes (2017) data loaded)=:#length(get_cheminfo())#Now let's resetreset_httk()# We should be back to our original number:num.chems == length(get_cheminfo())# Now add chemicals A, B, and C:my.new.data <- as.data.frame(c("A","B","C"),stringsAsFactors=FALSE)my.new.data <- cbind(my.new.data,as.data.frame(c(                     "111-11-2","222-22-0","333-33-5"),                     stringsAsFactors=FALSE))my.new.data <- cbind(my.new.data,as.data.frame(c("DTX1","DTX2","DTX3"),                    stringsAsFactors=FALSE))my.new.data <- cbind(my.new.data,as.data.frame(c(200,200,200)))my.new.data <- cbind(my.new.data,as.data.frame(c(2,3,4)))my.new.data <- cbind(my.new.data,as.data.frame(c(0.01,0.02,0.3)))my.new.data <- cbind(my.new.data,as.data.frame(c(0,10,100)))colnames(my.new.data) <- c("Name","CASRN","DTXSID","MW","LogP","Fup","CLint")chem.physical_and_invitro.data <- add_chemtable(my.new.data,                                  current.table=                                    chem.physical_and_invitro.data,                                  data.list=list(                                  Compound="Name",                                  CAS="CASRN",                                  DTXSID="DTXSID",                                  MW="MW",                                  logP="LogP",                                  Funbound.plasma="Fup",                                  Clint="CLint"),                                  species="Human",                                  reference="MyPaper 2015")parameterize_steadystate(chem.name="C")  calc_css(chem.name="B")                                # Initialize a column describing proton donors ("acids")my.new.data$pka.a <- NA # set chemical C to an acid (pKa_donor = 5):my.new.data[my.new.data$Name=="C","pka.a"] <- "5"chem.physical_and_invitro.data <- add_chemtable(my.new.data,                                  current.table=                                    chem.physical_and_invitro.data,                                 data.list=list(                                 Compound="Name",                                 CAS="CASRN",                                 DTXSID="DTXSID",                                 pKa_Donor="pka.a"),                                 species="Human",                                 reference="MyPaper 2015") # Note Rblood2plasma and hepatic bioavailability change (relative to above):parameterize_steadystate(chem.name="C")  # Initialize a column describing proton acceptors ("bases")my.new.data$pka.b <- NA # set chemical B to a base with multiple pka's (pKa_accept = 7 and 8):my.new.data[my.new.data$Name=="B","pka.b"] <- "7;8"chem.physical_and_invitro.data <- add_chemtable(my.new.data,                                  current.table=                                    chem.physical_and_invitro.data,                                 data.list=list(                                 Compound="Name",                                 CAS="CASRN",                                 DTXSID="DTXSID",                                 pKa_Accept="pka.b"),                                 species="Human",                                 reference="MyPaper 2015") # Note that average and max change (relative to above):calc_css(chem.name="B")

Draws ages from a smoothed distribution for a given gender/race combination

Description

This function should usually not be called directly by the user. It is used byhttkpop_generate() in "virtual-individuals" mode.

Usage

age_draw_smooth(gender, reth, nsamp, agelim_months, nhanes_mec_svy)

Arguments

Value

A named list with members 'ages_months' and 'ages_years', eachnumeric of lengthnsamp, giving the sampled ages in months and years.

Author(s)

Caroline Ring

References

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

Correct the measured intrinsive hepatic clearance for fraction free

Description

This function uses the free fraction estimated from Kilford et al. (2008) to increase the in vitro measure intrinsic hepatic clearance. The assumptionthat chemical that is bound in vitro is not available to be metabolized andtherefore the actual rate of clearance is actually faster. Note that in mosthigh throughput TK models included in the package this increase is offset bythe assumption of "restrictive clearance" – that is, the rate of hepaticmetabolism is slowed to account for the free fraction of chemical in plasma.This adjustment was made starting in Wetmore et al. (2015) in order to betterpredict plasma concentrations.

Usage

apply_clint_adjustment(  Clint,  Fu_hep = NULL,  Pow = NULL,  pKa_Donor = NULL,  pKa_Accept = NULL,  suppress.messages = FALSE)

Arguments

Value

Intrinsic hepatic clearance increased to take into account bindingin the in vitro assay

Author(s)

John Wambaugh

References

Kilford PJ, Gertz M, Houston JB, Galetin A (2008).“Hepatocellular binding of drugs: correction for unbound fraction in hepatocyte incubations using microsomal binding or drug lipophilicity data.”Drug Metabolism and Disposition,36(7), 1194–1197.doi:10.1124/dmd.108.020834. Wetmore BA, Wambaugh JF, Allen B, Ferguson SS, Sochaski MA, Setzer RW, Houck KA, Strope CL, Cantwell K, Judson RS, others (2015).“Incorporating high-throughput exposure predictions with dosimetry-adjusted in vitro bioactivity to inform chemical toxicity testing.”Toxicological Sciences,148(1), 121–136.doi:10.1093/toxsci/kfv171.

See Also

calc_hep_fu

Correct the measured fraction unbound in plasma for lipid binding

Description

This function uses the lipid binding correction estimated by Pearceet al. (2017) to decrease the fraction unbound in plasma(f_up). This correctionassumes that there is additional in vivo binding to lipid, whichhas a greater impact on neutral lipophilic compounds.

Usage

apply_fup_adjustment(  fup,  fup.correction = NULL,  Pow = NULL,  pKa_Donor = NULL,  pKa_Accept = NULL,  suppress.messages = FALSE,  minimum.Funbound.plasma = 1e-04)

Arguments

Value

Fraction unbound in plasma adjusted to take into account bindingin the in vitro assay

Author(s)

John Wambaugh

References

Kilford PJ, Gertz M, Houston JB, Galetin A (2008).“Hepatocellular binding of drugs: correction for unbound fraction in hepatocyte incubations using microsomal binding or drug lipophilicity data.”Drug Metabolism and Disposition,36(7), 1194–1197.doi:10.1124/dmd.108.020834. Wetmore BA, Wambaugh JF, Allen B, Ferguson SS, Sochaski MA, Setzer RW, Houck KA, Strope CL, Cantwell K, Judson RS, others (2015).“Incorporating high-throughput exposure predictions with dosimetry-adjusted in vitro bioactivity to inform chemical toxicity testing.”Toxicological Sciences,148(1), 121–136.doi:10.1093/toxsci/kfv171.

See Also

calc_fup_correction

Estimate well surface area

Description

Estimate geometry surface area of plastic in well plate based on well plateformat suggested values from Corning. option.plastic == TRUE (default) givenonzero surface area (sarea, m^2) option.bottom == TRUE (default) includessurface area of the bottom of the well in determining sarea. Optionallyinclude user values for working volume (v_working, m^3) and surface area.

Usage

armitage_estimate_sarea(  tcdata = NA,  user_assay_parameters = NA,  this.well_number = 384,  this.cell_yield = NA,  this.v_working = NA)

Arguments

Value

A data table composed of any input data.tabletcdatawith only the following columns either created or altered by this function:

Author(s)

Greg Honda, Meredith Scherer

References

Armitage JM, Wania F, Arnot JA (2014).“Application of mass balance models and the chemical activity concept to facilitate the use of in vitro toxicity data for risk assessment.”Environmental science & technology,48(16), 9770–9779.doi:10.1021/es501955g.

Armitage In Vitro Distribution Model

Description

Evaluate the Armitage model for chemical distributonin vitro. Takes inputas data table or vectors of values. Outputs a data table. Updates overthe model published in Armitage et al. (2014) include binding to plastic wallsand lipid and protein compartments in cells.

Usage

armitage_eval(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  casrn.vector = NA_character_,  nomconc.vector = 1,  this.well_number = 384,  this.FBSf = NA_real_,  tcdata = NA,  user_assay_parameters = NA,  this.sarea = NA_real_,  this.v_total = NA_real_,  this.v_working = NA_real_,  this.cell_yield = NA_real_,  this.Tsys = 37,  this.Tref = 298.15,  this.option.kbsa2 = FALSE,  this.option.swat2 = FALSE,  this.option.kpl2 = FALSE,  this.option.bottom = TRUE,  this.pseudooct = 0.01,  this.memblip = 0.04,  this.nlom = 0.05,  this.P_nlom = 0.035,  this.P_dom = 0.05,  this.P_cells = 1,  this.cell_pH = 7.4,  this.Anionic_VF = 0.175,  this.A_Prop_acid = 0.05,  this.A_Prop_base = 20,  this.Lyso_VF = 0.0068,  this.Lyso_Diam = 500,  this.Lyso_pH = 5.1,  this.csalt = 0.15,  this.celldensity = 1,  this.cellmass = 3,  this.f_oc = 1,  this.conc_ser_alb = 24,  this.conc_ser_lip = 1.9,  this.Vdom = 0,  this.pH = 7,  restrict.ion.partitioning = FALSE,  surface.area.switch = TRUE)

Arguments

Value

Author(s)

Greg Honda, Meredith Scherer adapted from code by James Armitage, Jon Arnot

References

Armitage JM, Wania F, Arnot JA (2014).“Application of mass balance models and the chemical activity concept to facilitate the use of in vitro toxicity data for risk assessment.”Environmental science & technology,48(16), 9770–9779.doi:10.1021/es501955g.

Honda GS, Pearce RG, Pham LL, Setzer RW, Wetmore BA, Sipes NS, Gilbert J, Franz B, Thomas RS, Wambaugh JF (2019).“Using the concordance of in vitro and in vivo data to evaluate extrapolation assumptions.”PloS one,14(5), e0217564.doi:10.1371/journal.pone.0217564.

Examples

library(httk)# Check to see if we have info on the chemical:"80-05-7" %in% get_cheminfo()#We do:temp <- armitage_eval(casrn.vector = c("80-05-7", "81-81-2"), this.FBSf = 0.1,this.well_number = 384, nomconc = 10)print(temp$cfree.invitro)# Check to see if we have info on the chemical:"793-24-8" %in% get_cheminfo()# Since we don't have any info, let's look up phys-chem from dashboard:cheminfo <- data.frame(  Compound="6-PPD",  CASRN="793-24-8",  DTXSID="DTXSID9025114",  logP=4.27,   logHenry=log10(7.69e-8),  logWSol=log10(1.58e-4),  MP=99.4,  MW=268.404  )  # Add the information to HTTK's database:chem.physical_and_invitro.data <- add_chemtable( cheminfo, current.table=chem.physical_and_invitro.data, data.list=list( Compound="Compound", CAS="CASRN",  DTXSID="DTXSID",  MW="MW",  logP="logP",  logHenry="logHenry",  logWSol="logWSol",  MP="MP"),  species="Human",  reference="CompTox Dashboard 31921")# Run the Armitage et al. (2014) model:out <- armitage_eval(  casrn.vector = "793-24-8",   this.FBSf = 0.1,  this.well_number = 384,   nomconc = 10)  print(out)

Add a parameter value to the chem.physical_and_invitro.data table

Description

This internal function is used byadd_chemtable to add a single new parameter to the table of chemical parameters. It should not be typicallyused from the command line.

Usage

augment.table(  this.table,  this.CAS,  compound.name = NULL,  this.property,  value,  species = NULL,  reference,  overwrite = FALSE,  sig.fig = 4,  clint.pvalue.overwrite = TRUE,  allow.na = FALSE,  suppress.messages = FALSE)

Arguments

Value

Author(s)

John Wambaugh

Find the best available ratio of the blood to plasma concentration constant.

Description

This function finds the best available constant ratio of the bloodconcentration to the plasma concentration, usingget_rblood2plasma andcalc_rblood2plasma.

Usage

available_rblood2plasma(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  species = "Human",  adjusted.Funbound.plasma = TRUE,  class.exclude = TRUE,  suppress.messages = FALSE)

Arguments

Details

Either retrieves a measured blood:plasma concentration ratio from thechem.physical_and_invitro.data table or calculates it using the red blood cellpartition coefficient predicted with Schmitt's method

If available, in vivo data (fromchem.physical_and_invitro.data) for thegiven species is returned, substituting the human in vivo value when missingfor other species. In the absence of in vivo data, the value is calculatedwithcalc_rblood2plasma for the given species. If Funbound.plasma isunvailable for the given species, the human Funbound.plasma is substituted.If none of these are available, the mean human Rblood2plasma fromchem.physical_and_invitro.data is returned. details than the description above ~~

Value

The blood to plasma chemical concentration ratio – measured if available,calculated if not.

Author(s)

Robert Pearce

See Also

calc_rblood2plasma

get_rblood2plasma

Examples

available_rblood2plasma(chem.name="Bisphenol A",adjusted.Funbound.plasma=FALSE)available_rblood2plasma(chem.name="Bisphenol A",species="Rat")

Assess the current performance of httk relative to historical benchmarks

Description

The function performs a series of "sanity checks" and predictive performancebenchmarks so that the impact of changes to the data, models, and implementation of the R package can be tested. Plots can be generated showing how the performance of the current version compares with past releases ofhttk.

Usage

benchmark_httk(  basic.check = TRUE,  calc_mc_css.check = TRUE,  in_vivo_stats.check = TRUE,  tissuepc.check = TRUE,  suppress.messages = TRUE,  make.plots = TRUE)

Arguments

Details

Historically some refinements made to one aspect of httk have unintentionallyimpacted other aspects. Most notably errors have occasionally been introduced with respect to units (v1.9, v2.1.0). This benchmarking tool is intended toreduce the chance of these errors occurring in the future.

Past performance was retroactively evaluated by manually installing previous versions of the package fromhttps://cran.r-project.org/src/contrib/Archive/httk/ andthen adding the code forbenchmark_httk at the command lineinterface.

The basic tests are important – if the output units for key functions are wrong, notmuch can be right. Past unit errors were linked to an incorrect unit conversions made within an individual function. Since the usage ofconvert_units became standard throughout httk,unit problems are hopefully less likely.

There are two Monte Carlo tests. One comparescalc_mc_css 95th percentilesteady-state plasma concentrations for a 1 mg/kg/day exposureagainst the Css values calculated by SimCyp and reported in Wetmore et al.(2012,2015). These have gradually diverged as the assumptions for httk haveshifted to better describe non-pharmaceutical, commercial chemicals.

The in vivo tests are in some ways the most important, as they establish theoverall predictability for httk for Cmax, AUC, and Css. The in vivo statistics are currently based on comparisons to the in vivodata compiled by Wambaugh et al. (2018). We see that when the tissuepartition coefficient calibrations were introduced in v1.6 that theoverall predictability for in vivo endpoints was reduced (increased RMSLE).If this phenomena continues as new in vivo evaluation data become available,we may need to revisit whether evaluation against experimentally-derived partition coefficients can actually be used for calibration, or just merelyfor establishing confidence intervals.

The partition coefficient tests provide an important check of the httkimplementation of the Schmitt (2008) model for tissue:plasma equilibrium distribution. These predictions heavily rely on accurate description of tissue composition and the ability to predict the ionization state of thecompounds being modeled.

Value

named list, whose elements depend on the selected checks

Author(s)

John Wambaugh

References

Davidson-Fritz SE, Ring CL, Evans MV, Schacht CM, Chang X, Breen M, Honda GS, Kenyon E, Linakis MW, Meade A, others (2025).“Enabling Transparent Toxicokinetic Modeling for Public Health Risk Assessment.”PLOS ONE,20(4), 1-40.doi:10.1371/journal.pone.0321321.

Find average blood masses by age.

Description

If blood mass fromblood_weight is negative or very small,then just default to the mean blood mass by age. (Geigy Scientific Tables,7th ed.)

Usage

blood_mass_correct(blood_mass, age_months, age_years, gender, weight)

Arguments

Value

A vector of blood masses in kg.

Author(s)

Caroline Ring

References

Geigy Pharmaceuticals, "Scientific Tables", 7th Edition, John Wiley and Sons (1970)

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

Predict blood mass.

Description

Predict blood mass based on body surface area and gender, using equationsfrom Bosgra et al. 2012

Usage

blood_weight(BSA, gender)

Arguments

Value

A vector of blood masses in kg the same length asBSA andgender.

Author(s)

Caroline Ring

References

Bosgra, Sieto, et al. "An improved model to predict physiologically based model parameters and their inter-individual variability from anthropometry." Critical reviews in toxicology 42.9 (2012): 751-767.

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

CDC BMI-for-age charts

Description

Charts giving the BMI-for-age percentiles for boys and girls ages 2-18

Usage

bmiage

Format

A data.table with 434 rows and 5 variables:

Sex

Female or Male

Agemos

Age in months

P5

The 5thpercentile BMI for the corresponding sex and age

P85

The 85thpercentile BMI for the corresponding sex and age

P95

The 95thpercentile BMI for the corresponding sex and age

Details

For children ages 2 to 18, weight class depends on the BMI-for-age percentile.

Underweight

<5th percentile

Normal weight

5th-85th percentile

Overweight

85th-95th percentile

Obese

>=95th percentile

Author(s)

Caroline Ring

Source

https://www.cdc.gov/growthcharts/data/zscore/bmiagerev.csv

References

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

Predict body surface area.

Description

Predict body surface area from weight, height, and age, using Mosteller'sformula for age>18 and Haycock's formula for age<18

Usage

body_surface_area(BW, H, age_years)

Arguments

Value

A vector of body surface areas in cm^2.

Author(s)

Caroline Ring

References

Mosteller, R. D. "Simplified calculation of body surface area." N Engl J Med 317 (1987): 1098..

Haycock, George B., George J. Schwartz, and David H. Wisotsky. "Geometric method for measuring body surface area: a height-weight formula validated in infants, children, and adults." The Journal of pediatrics 93.1 (1978): 62-66.

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

Predict bone mass

Description

Predict bone mass from age_years, height, weight, gender, using logisticequations fit to data from Baxter-Jones et al. 2011, or for infants < 1year, using equation from Koo et al. 2000 (See Price et al. 2003)

Usage

bone_mass_age(age_years, age_months, height, weight, gender)

Arguments

Value

Vector of bone masses.

Author(s)

Caroline Ring

References

Baxter-Jones, Adam DG, et al. "Bone mineral accrual from 8 to 30 years of age: an estimation of peak bone mass." Journal of Bone and Mineral Research 26.8 (2011): 1729-1739.

Koo, Winston WK, and Elaine M. Hockman. "Physiologic predictors of lumbar spine bone mass in neonates." Pediatric research 48.4 (2000): 485-489.

Price, Paul S., et al. "Modeling interindividual variation in physiological factors used in PBPK models of humans." Critical reviews in toxicology 33.5 (2003): 469-503.

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

Predict brain mass.

Description

Predict brain mass from gender and age.

Usage

brain_mass(gender, age_years)

Arguments

Value

A vector of brain masses in kg.

Author(s)

Caroline Ring

References

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

Calculate the analytic steady state plasma concentration.

Description

This function calculates the analytic steady state plasma or venous blood concentrations as a result of infusion dosing for the three compartment and multiple compartment PBTK models.

Usage

calc_analytic_css(  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  parameters = NULL,  species = "human",  daily.dose = NULL,  dose = 1,  dose.units = "mg/kg/day",  route = "oral",  output.units = "uM",  model = "pbtk",  concentration = "plasma",  suppress.messages = FALSE,  tissue = NULL,  bioactive.free.invivo = FALSE,  IVIVE = NULL,  parameterize.args.list = list(),  ...)

Arguments

Details

Concentrations are calculated for the specifed model with constant oral infusion dosing. All tissues other than gut, liver, and lung are the product of the steady state plasma concentration and the tissue to plasma partition coefficient.

Only four sets of IVIVE assumptions that performed well in Honda et al. (2019) are currently included inhonda.ivive:"Honda1" through "Honda4". The use of max (peak) concentration can not be currently be calculated withcalc_analytic_css. The httk default settings correspond to "Honda3":

"Honda1" uses plasma concentration, restrictive clearance, and treats the unbound invivo concentration as bioactive. For IVIVE, any input nominal concentration in vitro should be converted to cfree.invitro usingarmitage_eval, otherwise performance will be the same as "Honda2".

Value

Steady state plasma concentration in specified units

Author(s)

Robert Pearce, John Wambaugh, Greg Honda, Miyuki Breen

References

Honda GS, Pearce RG, Pham LL, Setzer RW, Wetmore BA, Sipes NS, Gilbert J, Franz B, Thomas RS, Wambaugh JF (2019).“Using the concordance of in vitro and in vivo data to evaluate extrapolation assumptions.”PloS one,14(5), e0217564.doi:10.1371/journal.pone.0217564.

See Also

calc_css

Examples

calc_analytic_css(chem.name='Bisphenol-A',output.units='mg/L',                 model='3compartment',concentration='blood')# Test that the underlying PK models give the same answers:calc_analytic_css(chem.cas="15972-60-8")calc_analytic_css(chem.cas="15972-60-8",model="1compartment")calc_analytic_css(chem.cas="15972-60-8",model="pbtk")calc_analytic_css(chem.cas="15972-60-8",model="3compartment")calc_analytic_css(chem.name='Bisphenol-A',tissue='liver',species='rabbit',                 parameterize.args.list = list(                                default.to.human=TRUE,                                adjusted.Funbound.plasma=TRUE,                                regression=TRUE,                                minimum.Funbound.plasma=1e-4),daily.dose=2)calc_analytic_css(chem.name="bisphenol a",model="1compartment")calc_analytic_css(chem.cas="80-05-7",model="3compartmentss")params <- parameterize_pbtk(chem.cas="80-05-7") calc_analytic_css(parameters=params,model="pbtk")# Try various chemicals with differing parameter sources/issues:calc_analytic_css(chem.name="Betaxolol")calc_analytic_css(chem.name="Tacrine",model="pbtk")calc_analytic_css(chem.name="Dicofol",model="1compartment")calc_analytic_css(chem.name="Diflubenzuron",model="3compartment")calc_analytic_css(chem.name="Theobromine",model="3compartmentss")# permutations on steady-state for the 1compartment modelcalc_analytic_css(chem.name="bisphenol a",                  model="1compartment")calc_analytic_css(chem.cas="80-05-7",                  model="1compartment")calc_analytic_css(parameters=parameterize_1comp(chem.cas="80-05-7"),                  model="1compartment")calc_analytic_css(chem.cas="80-05-7",                  model="1compartment",                  tissue="liver")calc_analytic_css(chem.cas="80-05-7",                  model="1compartment",                  tissue="brain")# permutations on steady-state for the 3compartment modelcalc_analytic_css(chem.cas="80-05-7",                  model="3compartment")calc_analytic_css(parameters=parameterize_3comp(chem.cas="80-05-7"),                  model="3compartment")calc_analytic_css(chem.name="bisphenol a",                  model="3compartment",                  tissue="liver")calc_analytic_css(chem.name="bisphenol a",                  model="3compartment",                  tissue="brain")# permurtations on steady-state for the pbtk model:calc_analytic_css(chem.cas="80-05-7",                  model="pbtk")calc_analytic_css(parameters=parameterize_pbtk(chem.cas="80-05-7"),                  model="pbtk")calc_analytic_css(chem.name="bisphenol a",                  model="pbtk",                  tissue="liver")calc_analytic_css(chem.name="bisphenol a",                  model="pbtk",                  tissue="brain")# Test oral absorption functionality:# By default we now include calculation of Fabs and Fgut (always had Fhep):calc_analytic_css(chem.name="bisphenol a",                  model="pbtk")# Therefore if we set Fabs = Fgut = 1 with keetit100=TRUE, we should get a# higher predicted plasma steady-state concentration:calc_analytic_css(chem.name="bisphenol a",                  model="pbtk",                  Caco2.options=list(keepit100=TRUE))

Calculate the analytic steady state concentration for the one compartment model.

Description

This function calculates the analytic steady state plasma or venous blood concentrations as a result of infusion dosing.

Usage

calc_analytic_css_1comp(  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  species = "Human",  parameters = NULL,  dosing = list(daily.dose = 1),  hourly.dose = NULL,  dose.units = "mg",  concentration = "plasma",  suppress.messages = FALSE,  recalc.blood2plasma = FALSE,  tissue = NULL,  restrictive.clearance = TRUE,  bioactive.free.invivo = FALSE,  Caco2.options = list(),  ...)

Arguments

Value

Steady state plasma concentration in mg/L units

Author(s)

Robert Pearce and John Wambaugh

See Also

calc_analytic_css

parameterize_1comp

Calculate the analytic steady state concentration for model 3compartment

Description

This function calculates the analytic steady state plasma or blood concentrations as a result of constant oral infusion dosing.The three compartment model (Pearce et al. 2017)describes the amount of chemical inthree key tissues of the body: the liver, the portal vein (essentially, oral absorptionfrom the gut), and a systemic compartment ("sc") representing the rest of the body.Seesolve_3comp for additional details. The analyticalsteady-state solution for the the three compartment model is:

C^{ss}_{plasma} = \frac{dose}{f_{up}*Q_{GFR} + Cl_{h} + \frac{Cl_{h}}{Q_{l}}\frac{f_{up}}{R_{b:p}}Q_{GFR}}

C^{ss}_{blood} = R_{b:p}*C^{ss}_{plasma}

where Q_GFR is the glomerular filtrationrate in the kidney, Q_l is the total liver blood flow (hepatic artery plustotal vein),Cl_h is the chemical-specific whole liver metabolism clearance (scaled up from intrinsic clearance, which does not depend on flow),f_up is the chemical-specific fraction unbound in plasma, R_b:p is the chemical specific ratio of concentrations in blood:plasma.

Usage

calc_analytic_css_3comp(  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  species = "Human",  parameters = NULL,  dosing = list(daily.dose = 1),  hourly.dose = NULL,  dose.units = "mg",  concentration = "plasma",  suppress.messages = FALSE,  recalc.blood2plasma = FALSE,  tissue = NULL,  restrictive.clearance = TRUE,  bioactive.free.invivo = FALSE,  Caco2.options = list(),  ...)

Arguments

Value

Steady state plasma concentration in mg/L units

Author(s)

Robert Pearce and John Wambaugh

References

Pearce RG, Setzer RW, Strope CL, Wambaugh JF, Sipes NS (2017).“Httk: R package for high-throughput toxicokinetics.”Journal of Statistical Software,79(4), 1.doi:10.18637/jss.v079.i04.

See Also

calc_analytic_css

parameterize_3comp

Calculate the analytic steady state concentration for model 3compartment

Description

This function calculates the analytic steady state plasma or blood concentrations as a result of constant oral infusion dosing.The three compartment model (Pearce et al. 2017)describes the amount of chemical inthree key tissues of the body: the liver, the portal vein (essentially, oral absorptionfrom the gut), and a systemic compartment ("sc") representing the rest of the body.Seesolve_3comp for additional details. The analyticalsteady-state solution for the the three compartment model is:

C^{ss}_{plasma} = \frac{dose}{f_{up}*Q_{GFR} + Cl_{h} + \frac{Cl_{h}}{Q_{l}}\frac{f_{up}}{R_{b:p}}Q_{GFR}}

C^{ss}_{blood} = R_{b:p}*C^{ss}_{plasma}

where Q_GFR is the glomerular filtrationrate in the kidney, Q_l is the total liver blood flow (hepatic artery plustotal vein),Cl_h is the chemical-specific whole liver metabolism clearance (scaled up from intrinsic clearance, which does not depend on flow),f_up is the chemical-specific fraction unbound in plasma, R_b:p is the chemical specific ratio of concentrations in blood:plasma.

Usage

calc_analytic_css_3comp2(  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  species = "Human",  parameters = NULL,  dosing = list(daily.dose = 1),  hourly.dose = NULL,  dose.units = "mg",  concentration = "plasma",  suppress.messages = FALSE,  recalc.blood2plasma = FALSE,  tissue = NULL,  route = "oral",  restrictive.clearance = TRUE,  bioactive.free.invivo = FALSE,  Caco2.options = list(),  exhalation = TRUE,  ...)

Arguments

Value

Steady state plasma concentration in mg/L units

Author(s)

Robert Pearce and John Wambaugh

References

Pearce RG, Setzer RW, Strope CL, Wambaugh JF, Sipes NS (2017).“Httk: R package for high-throughput toxicokinetics.”Journal of Statistical Software,79(4), 1.doi:10.18637/jss.v079.i04.

See Also

calc_analytic_css

parameterize_3comp

Calculate the analytic steady state concentration for the three compartmentsteady-state model

Description

This function calculates the steady state plasma or venous blood concentrations as a result of constant oral infusion dosing. The equation, initally used for high throughput in vitro-in vivo extrapolation in(Rotroff et al. 2010) and later given in (Wetmore et al. 2012), assumes that the concentration is the inverse of the total clearance, which is the sum of hepatic metabolismand renal filatrion:

C^{ss}_{plasma} = \frac{dose}{f_{up}*Q_{GFR}+Cl_{h}}

C^{ss}_{blood} = R_{b:p}*C^{ss}_{plasma}

where Q_GFR is the glomerular filtrationrate in the kidney, Cl_h is the chemical-specific whole liver metabolism clearance (scaled up from intrinsic clearance, which does not depend on flow),f_up is the chemical-specific fraction unbound in plasma, R_b:p is the chemical specific ratio of concentrations in blood:plasma.

Usage

calc_analytic_css_3compss(  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  parameters = NULL,  dosing = list(daily.dose = 1),  hourly.dose = NULL,  dose.units = "mg",  concentration = "plasma",  suppress.messages = FALSE,  recalc.blood2plasma = FALSE,  tissue = NULL,  restrictive.clearance = TRUE,  bioactive.free.invivo = FALSE,  Caco2.options = list(),  ...)

Arguments

Details

This equation is a simplification of the steady-state plasma concentrationin the three-comprtment model (seesolve_3comp), neglecting ahigher order term that causes this Css to be higher for very rapidly clearedchemicals.

Value

Steady state plasma concentration in mg/L units

Author(s)

Robert Pearce and John Wambaugh

References

Rotroff DM, Wetmore BA, Dix DJ, Ferguson SS, Clewell HJ, Houck KA, LeCluyse EL, Andersen ME, Judson RS, Smith CM, others (2010).“Incorporating human dosimetry and exposure into high-throughput in vitro toxicity screening.”Toxicological Sciences,117(2), 348–358.doi:10.1093/toxsci/kfq220.

Wetmore BA, Wambaugh JF, Ferguson SS, Sochaski MA, Rotroff DM, Freeman K, Clewell III HJ, Dix DJ, Andersen ME, Houck KA, others (2012).“Integration of dosimetry, exposure, and high-throughput screening data in chemical toxicity assessment.”Toxicological Sciences,125(1), 157–174.doi:10.1093/toxsci/kfr254.

See Also

calc_analytic_css

parameterize_steadystate

Calculate the analytic steady state plasma concentration for model pbtk.

Description

This function calculates the analytic steady state concentration (mg/L) as a resultof constant oral infusion dosing. Concentrations are returned for plasma by default, but varioustissues or blood concentrations can also be given as specified.

Usage

calc_analytic_css_pbtk(  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  species = "Human",  parameters = NULL,  dosing = list(daily.dose = 1),  hourly.dose = NULL,  dose.units = "mg",  concentration = "plasma",  suppress.messages = FALSE,  recalc.blood2plasma = FALSE,  tissue = NULL,  restrictive.clearance = TRUE,  bioactive.free.invivo = FALSE,  Caco2.options = list(),  ...)

Arguments

Details

The PBTK model (Pearce et al. 2017) predicts the amount of chemical invarious tissues of the body. A system of ordinarydifferential equations describes how the amounts in each tissue change as a function of time. The analytic steady-state equation was found by algebraically solving for the tissue concentrations that result in eachequation being zero – thus determining the concentration at which there is no changeover time as the result of a fixed infusion dose rate.

The analytical solution is:

C^{ss}_{ven} = \frac{dose rate * \frac{Q_{liver} + Q_{gut}}{\frac{f_{up}}{R_{b:p}}*Cl_{metabolism} + (Q_{liver}+Q_{gut})}}{Q_{cardiac} - \frac{(Q_{liver} + Q_{gut})^2}{\frac{f_{up}}{R_{b:p}}*Cl_{metabolism} + (Q_{liver}+Q_{gut})} - \frac{(Q_{kidney})^2}{\frac{f_{up}}{R_{b:p}}*Q_{GFR}+Q_{kideny}}-Q_{rest}}

C^{ss}_{plasma} = \frac{C^{ss}_{ven}}{R_{b:p}}

C^{ss}_{tissue} = \frac{K_{tissue:fuplasma}*f_{up}}{R_{b:p}}*C^{ss}_{ven}

where Q_cardiac is the cardiac output, Q_gfr is the glomerular filtrationrate in the kidney, other Q's indicate blood flows to various tissues, Cl_metabolism is the chemical-specific whole liver metabolism clearance,f_up is the chemical-specific fraction unbound in plasma, R_b2p is the chemical specific ratio of concentrations in blood:plasma, K_tissue2fuplasmais the chemical- and tissue-specific equilibrium partition coefficientand dose rate has units of mg/kg/day.

Value

Steady state plasma concentration in mg/L units

Author(s)

Robert Pearce and John Wambaugh

References

Pearce RG, Setzer RW, Strope CL, Wambaugh JF, Sipes NS (2017).“Httk: R package for high-throughput toxicokinetics.”Journal of Statistical Software,79(4), 1.doi:10.18637/jss.v079.i04.

See Also

calc_analytic_css

parameterize_pbtk

Calculate the steady state concentration for the sum of clearancessteady-state model with exhalation

Description

This function calculates the analytic steady state plasma or venous blood concentrations as a result of infusion dosing.

Usage

calc_analytic_css_sumclearances(  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  species = "Human",  parameters = NULL,  dosing = list(daily.dose = 1),  hourly.dose = NULL,  dose.units = "mg",  concentration = "plasma",  Caco2.options = NULL,  suppress.messages = FALSE,  recalc.blood2plasma = FALSE,  tissue = NULL,  route = "oral",  restrictive.clearance = TRUE,  bioactive.free.invivo = FALSE,  ...)

Arguments

Value

Steady state plasma concentration in mg/L units

Author(s)

John Wambaugh

See Also

calc_analytic_css

parameterize_steadystate

Calculate the fractional contributions to total clearance

Description

Steady-state clearance is a function of multiple processes. For example,meabolism in the liver and glomerular filtration in the kidney. This functiontakes a list of parameters potentially impacting total clearance anditeratively sets all but one of the paramters to zero. This allowscalculation of the fration of total clearance driven by that parameter.

Usage

calc_clearance_frac(  fraction.params = c("Clint", "Qgfrc"),  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  species = "human",  default.to.human = FALSE,  suppress.messages = FALSE,  model = "3compartmentss",  restrictive.clearance = TRUE,  parameterize.args = list(),  analytic_css.args = list())

Arguments

Value

A numeric fraction unpbound in plasma between zero and one

Author(s)

John Wambaugh

Examples

# 3compartmentss model:calc_clearance_frac(chem.name="bisphenola")# pbtk model:calc_clearance_frac(chem.name="bisphenola",                    model="pbtk",                    fraction.params=c("Qgfrc","Clmetabolismc"))# A model with exhalation:# sumclearances model:calc_clearance_frac(chem.name="bisphenola",                    model="sumclearances",                    fraction.params=c("Clint","Qgfrc","Qalvc"))calc_clearance_frac(chem.name="toluene",                    model="sumclearances",                    fraction.params=c("Clint","Qgfrc","Qalvc"))                      # 3comp2 model:calc_clearance_frac(chem.name="toluene",                    model="3compartment2",                    fraction.params=c("Clmetabolismc","Qgfrc","Qalvc"))

Find the steady state concentration and the day it is reached.

Description

This function finds the day a chemical comes within the specifiedrange of the analytical steady state venous blood or plasmaconcentration(from calc_analytic_css) for the multiple compartment, threecompartment, and one compartment models, the fraction of the true steadystate value reached on that day, the maximum concentration, and the averageconcentration at the end of the simulation.

Usage

calc_css(  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  parameters = NULL,  species = "Human",  f = 0.01,  daily.dose = 1,  doses.per.day = 3,  dose.units = "mg/kg",  route = "oral",  days = 21,  output.units = "uM",  suppress.messages = FALSE,  tissue = NULL,  model = "pbtk",  f.change = 1e-05,  dosing = NULL,  parameterize.args.list = list(),  ...)

Arguments

Value

Author(s)

Robert Pearce, John Wambaugh

See Also

calc_analytic_css

Examples

calc_css(chem.name='Bisphenol-A',doses.per.day=5,f=.001,output.units='mg/L')parms <- parameterize_3comp(chem.name='Bisphenol-A')parms$Funbound.plasma <- .07calc_css(chem.name='Bisphenol-A',parameters=parms,model='3compartment')out <- solve_pbtk(chem.name = "Bisphenol A",  days = 50,  daily.dose=1,  doses.per.day = 3)plot.data <- as.data.frame(out)css <- calc_analytic_css(chem.name = "Bisphenol A")library("ggplot2")c.vs.t <- ggplot(plot.data,aes(time, Cplasma)) + geom_line() +geom_hline(yintercept = css) + ylab("Plasma Concentration (uM)") +xlab("Day") + theme(axis.text = element_text(size = 16), axis.title =element_text(size = 16), plot.title = element_text(size = 17)) +ggtitle("Bisphenol A")print(c.vs.t)calc_css(chem.name='nicotine', model="1compartment")calc_css(chem.name='nicotine', model="3compartment")calc_css(chem.name="endrin")

Calculate Dermal Equivalent Dose

Description

This functions converts a steady state plasma concetration for a givendermal exposure scenario to an equivalent steady state media concentrationfor a single dose.

Usage

calc_dermal_equiv(  conc,  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  parameters = NULL,  days = 20,  doses.per.day = 3,  skin_depth = 0.3,  skin.pH = 7,  Vmedia = 0.001,  Fskinexposed = 0.1,  ...)

Arguments

Details

Returned dose is dependent on doses.per.day.

Value

Equivalent dose in specified units, default of mg/kg BW/day.

Author(s)

Annabel Meade

Calculate the distribution coefficient

Description

This function estimates the ratio of the equilibrium concentrations ofa compound in octanol and water, taking into account the charge of thecompound. Given the pH, we assume the neutral (uncharged) fraction ofcompound partitions according to the hydrophobicity (P_ow). We assume thatonly a fraction alpha (defaults to 0.001 – Schmitt (2008)) of the chargedcompound partitions into lipid (octanol):D_ow = P_ow*(F_neutral + alpha*F_charged)Fractions charged are calculatedaccording to hydrogen ionization equilibria (pKa_Donor, pKa_Accept) usingcalc_ionization.

Usage

calc_dow(  Pow = NULL,  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  pH = NULL,  pKa_Donor = NULL,  pKa_Accept = NULL,  fraction_charged = NULL,  alpha = 0.001)

Arguments

Value

Distribution coefficient (numeric)

Author(s)

Robert Pearce and John Wambaugh

References

Schmitt W (2008).“General approach for the calculation of tissue to plasma partition coefficients.”Toxicology in vitro,22(2), 457–467.doi:10.1016/j.tiv.2007.09.010.

Pearce RG, Setzer RW, Davis JL, Wambaugh JF (2017).“Evaluation and calibration of high-throughput predictions of chemical distribution to tissues.”Journal of pharmacokinetics and pharmacodynamics,44, 549–565.doi:10.1007/s10928-017-9548-7.

Strope CL, Mansouri K, Clewell III HJ, Rabinowitz JR, Stevens C, Wambaugh JF (2018).“High-throughput in-silico prediction of ionization equilibria for pharmacokinetic modeling.”Science of The Total Environment,615, 150–160.doi:10.1016/j.scitotenv.2017.09.033.

See Also

calc_ionization

Calculate the elimination rate for a one compartment model

Description

This function calculates an elimination rate from the three compartmentsteady state model where elimination is entirely due to metablism by theliver and glomerular filtration in the kidneys.

Usage

calc_elimination_rate(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  species = "Human",  model = "3compartmentss",  suppress.messages = TRUE,  ...)

Arguments

Details

Elimination rate calculated by dividing the total clearance (using thedefault -stirred hepatic model) by the volume of distribution. Whenspecies is specified as rabbit, dog, or mouse, the function uses theappropriate physiological data(volumes and flows) but substitues humanfraction unbound, partition coefficients, and intrinsic hepatic clearance.

Value

Author(s)

John Wambaugh

References

Schmitt W (2008).“General approach for the calculation of tissue to plasma partition coefficients.”Toxicology in vitro,22(2), 457–467.doi:10.1016/j.tiv.2007.09.010.

Dawson DE, Ingle BL, Phillips KA, Nichols JW, Wambaugh JF, Tornero-Velez R (2021).“Designing QSARs for Parameters of High-Throughput Toxicokinetic Models Using Open-Source Descriptors.”Environmental Science & Technology,55(9), 6505-6517.doi:10.1021/acs.est.0c06117.PMID: 33856768, https://doi.org/10.1021/acs.est.0c06117.

Kilford PJ, Gertz M, Houston JB, Galetin A (2008).“Hepatocellular binding of drugs: correction for unbound fraction in hepatocyte incubations using microsomal binding or drug lipophilicity data.”Drug Metabolism and Disposition,36(7), 1194–1197.doi:10.1124/dmd.108.020834.

See Also

calc_total_clearance for calculation of total clearance

calc_vdist for calculation of volume of distribution

Examples

calc_elimination_rate(chem.name="Bisphenol A")## Not run: calc_elimination_rate(chem.name="Bisphenol A",species="Rat")# non-restrictive clearance should be faster:kelim1 <- calc_elimination_rate(chem.cas="80-05-7")kelim2 <- calc_elimination_rate(chem.cas="80-05-7",                                 restrictive.clearance=FALSE)if (!(kelim2 > kelim1)) stop("kelim2 is not faster than kelim1")## End(Not run)

Functions for calculating the bioavaialble fractions from oral doses

Description

These functions calculate the fraction of chemical absorbed from the gutbased upon in vitro measured Caco-2 membrane permeability data.Caco-2 permeabilities (10^{-6} cm/s) are related to effective permeability based on Yang et al. (2007).These functions calculate the fraction absorbed (calc_fabs.oral – S Darwich et al. (2010) andYu and Amidon (1999)), the fractionsurviving first pass gut metabolism (calc_fgut.oral), and the overall systemicoral bioavailability(calc_fbio.oral). Note that the first pass hepatic clearance is calculated within theparameterization and other functions. usingcalc_hep_bioavailability Absorption rate is calculated according to Fick's law (LennernÄs (1997)) assuming low blood concentrations.

Usage

calc_fbio.oral(  parameters = NULL,  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  species = "Human",  suppress.messages = FALSE,  ...)calc_fabs.oral(  parameters = NULL,  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  species = "Human",  suppress.messages = FALSE,  Caco2.Pab.default = 1.6)calc_peff(  parameters = NULL,  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  species = "Human",  suppress.messages = FALSE,  Caco2.Pab = NULL,  parameterize.args.list = list())calc_kgutabs(  parameters = NULL,  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  species = "Human",  suppress.messages = FALSE,  parameterize.args.list = list())calc_fgut.oral(  parameters = NULL,  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  species = "Human",  suppress.messages = FALSE,  Caco2.Pab.default = 1.6,  parameterize.args.list = list())

Arguments

Details

We assume that systemic oral bioavailability (F_{bio})consists of three components: (1) the fraction of chemical absorbed from intestinal lumen into enterocytes (F_{abs}), (2) the fraction surviving intestinal metabolism (F_{gut}), and (3) the fraction surviving first-pass hepatic metabolism (F_{hep}). This function returns (F_{abs}*F_{gut}).

We model systemic oral bioavailability asF_{bio}=F_{abs}*F_{gut}*F_{hep}.F_{hep} is estimated from in vitro TK data usingcalc_hep_bioavailability.IfF_{bio} has been measured in vivo and is found intablechem.physical_and_invitro.data then we setF_{abs}*F_{gut} to the measured value divided byF_{hep}.Otherwise, if Caco2 membrane permeability data or predictionsare availableF_{abs} is estimated usingcalc_fgut.oral.Intrinsic hepatic metabolism is used to very roughly estimate (F_{gut})usingcalc_fgut.oral.If argument keepit100 is used then there is complete absorption from the gut(that is,F_{abs}=F_{gut}=1).

Value

Functions

• calc_fabs.oral(): Calculate the fraction absorbed in the gut (Darwich et al., 2010)

• calc_peff(): Calculate the effective gut permeability rate (10^-4 cm/s)

• calc_kgutabs(): Calculate the gut absorption rate (1/h)

• calc_fgut.oral(): Calculate the fraction of chemical surviving first pass metabolism in the gut

Author(s)

Gregory Honda and John Wambaugh

References

S Darwich A, Neuhoff S, Jamei M, Rostami-Hodjegan A (2010).“Interplay of metabolism and transport in determining oral drug absorption and gut wall metabolism: a simulation assessment using the 'Advanced Dissolution, Absorption, Metabolism (ADAM)' model.”Current drug metabolism,11(9), 716–729.doi:10.2174/138920010794328913.

Yang J, Jamei M, Yeo KR, Tucker GT, Rostami-Hodjegan A (2007).“Prediction of intestinal first-pass drug metabolism.”Current drug metabolism,8(7), 676–684.doi:10.2174/138920007782109733.

Honda GS, Kenyon EM, Davidson-Fritz S, Dinallo R, El Masri H, Korol-Bexell E, Li L, Angus D, Pearce RG, Sayre RR, others (2025).“Impact of gut permeability on estimation of oral bioavailability for chemicals in commerce and the environment.”ALTEX-Alternatives to animal experimentation,42(1), 56–74.doi:10.14573/altex.2403271.

Yu LX, Amidon GL (1999).“A compartmental absorption and transit model for estimating oral drug absorption.”International journal of pharmaceutics,186(2), 119–125.doi:10.1016/S0378-5173(99)00147-7.

LennernÄs H (1997).“Human jejunal effective permeability and its correlation with preclinical drug absorption models.”Journal of Pharmacy and Pharmacology,49(7), 627–638.doi:10.1111/j.2042-7158.1997.tb06084.x.

Calculate maternal-fetal physiological parameters

Description

This function uses the equations from Kapraun (2019) to calculate chemical-independent physiological paramreters as a function of gestational age in weeks.

Usage

calc_fetal_phys(week = 12, ...)

Arguments

Details

BW = pre_pregnant_BW + BW_cubic_theta1 * tw + BW_cubic_theta2 * tw^2 + BW_cubic_theta3 * tw^3

Wadipose = Wadipose_linear_theta0 + Wadipose_linear_theta1 * tw ;

Wfkidney = 0.001 * Wfkidney_gompertz_theta0 * exp ( Wfkidney_gompertz_theta1 / Wfkidney_gompertz_theta2 * ( 1 - exp ( - Wfkidney_gompertz_theta2 * tw ) ) ) ;

Wfthyroid = 0.001 * Wfthyroid_gompertz_theta0 * exp ( Wfthyroid_gompertz_theta1 / Wfthyroid_gompertz_theta2 * ( 1 - exp ( - Wfthyroid_gompertz_theta2 * tw ) ) ) ;

Wfliver = 0.001 * Wfliver_gompertz_theta0 * exp ( Wfliver_gompertz_theta1 / Wfliver_gompertz_theta2 * ( 1 - exp ( - Wfliver_gompertz_theta2 * tw ) ) ) ;

Wfbrain = 0.001 * Wfbrain_gompertz_theta0 * exp ( Wfbrain_gompertz_theta1 / Wfbrain_gompertz_theta2 * ( 1 - exp ( - Wfbrain_gompertz_theta2 * tw ) ) ) ;

Wfgut = 0.001 * Wfgut_gompertz_theta0 * exp ( Wfgut_gompertz_theta1 / Wfgut_gompertz_theta2 * ( 1 - exp ( - Wfgut_gompertz_theta2 * tw ) ) ) ;

Wflung = 0.001 * Wflung_gompertz_theta0 * exp ( Wflung_gompertz_theta1 / Wflung_gompertz_theta2 * ( 1 - exp ( - Wflung_gompertz_theta2 * tw ) ) ) ;

hematocrit = ( hematocrit_quadratic_theta0 + hematocrit_quadratic_theta1 * tw + hematocrit_quadratic_theta2 * pow ( tw , 2 ) ) / 100 ;

Rblood2plasma = 1 - hematocrit + hematocrit * Krbc2pu * Fraction_unbound_plasma ;

fhematocrit = ( fhematocrit_cubic_theta1 * tw + fhematocrit_cubic_theta2 * pow ( tw , 2 ) + fhematocrit_cubic_theta3 * pow ( tw , 3 ) ) / 100 ;

Rfblood2plasma = 1 - fhematocrit + fhematocrit * Kfrbc2pu * Fraction_unbound_plasma_fetus ;

fBW = 0.001 * fBW_gompertz_theta0 * exp ( fBW_gompertz_theta1 / fBW_gompertz_theta2 * ( 1 - exp ( - fBW_gompertz_theta2 * tw ) ) ) ;

Vplacenta = 0.001 * ( Vplacenta_cubic_theta1 * tw + Vplacenta_cubic_theta2 * pow ( tw , 2 ) + Vplacenta_cubic_theta3 * pow ( tw , 3 ) ) ;

Vamnf = 0.001 * Vamnf_logistic_theta0 / ( 1 + exp ( - Vamnf_logistic_theta1 * ( tw - Vamnf_logistic_theta2 ) ) ) ;

Vplasma = Vplasma_mod_logistic_theta0 / ( 1 + exp ( - Vplasma_mod_logistic_theta1 * ( tw - Vplasma_mod_logistic_theta2 ) ) ) + Vplasma_mod_logistic_theta3 ;

Vrbcs = hematocrit / ( 1 - hematocrit ) * Vplasma ;

Vven = venous_blood_fraction * ( Vrbcs + Vplasma ) ;

Vart = arterial_blood_fraction * ( Vrbcs + Vplasma ) ;

Vadipose = 1 / adipose_density * Wadipose ;

Vffmx = 1 / ffmx_density * ( BW - Wadipose - ( fBW + placenta_density * Vplacenta + amnf_density * Vamnf ) ) ;

Vallx = Vart + Vven + Vthyroid + Vkidney + Vgut + Vliver + Vlung ;

Vrest = Vffmx - Vallx ;

Vfart = 0.001 * arterial_blood_fraction * fblood_weight_ratio * fBW ;

Vfven = 0.001 * venous_blood_fraction * fblood_weight_ratio * fBW ;

Vfkidney = 1 / kidney_density * Wfkidney ;

Vfthyroid = 1 / thyroid_density * Wfthyroid ;

Vfliver = 1 / liver_density * Wfliver ;

Vfbrain = 1 / brain_density * Wfbrain ;

Vfgut = 1 / gut_density * Wfgut ;

Vflung = 1 / lung_density * Wflung ;

Vfrest = fBW - ( Vfart + Vfven + Vfbrain + Vfkidney + Vfthyroid + Vfliver + Vfgut + Vflung ) ;

Qcardiac = 24 * ( Qcardiac_cubic_theta0 + Qcardiac_cubic_theta1 * tw + Qcardiac_cubic_theta2 * pow ( tw , 2 ) + Qcardiac_cubic_theta3 * pow ( tw , 3 ) ) ;

Qgut = 0.01 * ( Qgut_percent_initial + ( Qgut_percent_terminal - Qgut_percent_initial ) / term * tw ) * Qcardiac ;

Qkidney = 24 * ( Qkidney_cubic_theta0 + Qkidney_cubic_theta1 * tw + Qkidney_cubic_theta2 * pow ( tw , 2 ) + Qkidney_cubic_theta3 * pow ( tw , 3 ) ) ;

Qliver = 0.01 * ( Qliver_percent_initial + ( Qliver_percent_terminal - Qliver_percent_initial ) / term * tw ) * Qcardiac ;

Qthyroid = 0.01 * ( Qthyroid_percent_initial + ( Qthyroid_percent_terminal - Qthyroid_percent_terminal ) / term * tw ) * Qcardiac ;

Qplacenta = 24 * Qplacenta_linear_theta1 * 1000 * Vplacenta ;

Qadipose = 0.01 * ( Qadipose_percent_initial + ( Qadipose_percent_terminal - Qadipose_percent_initial ) / term * tw ) * Qcardiac ;

Qrest = Qcardiac - ( Qgut + Qkidney + Qliver + Qthyroid + Qplacenta + Qadipose ) ;

Qgfr = 60 * 24 * 0.001 * ( Qgfr_quadratic_theta0 + Qgfr_quadratic_theta1 * tw + Qgfr_quadratic_theta2 * pow ( tw , 2 ) ) ;

Qfrvtl = 60 * 24 * 0.001 * Qfrvtl_logistic_theta0 / ( 1 + exp ( - Qfrvtl_logistic_theta1 * ( tw - Qfrvtl_logistic_theta2 ) ) ) ;

Qflvtl = 60 * 24 * 0.001 * Qflvtl_logistic_theta0 / ( 1 + exp ( - Qflvtl_logistic_theta1 * ( tw - Qflvtl_logistic_theta2 ) ) ) ;

Qfda = 60 * 24 * 0.001 * Qfda_logistic_theta0 / ( 1 + exp ( - Qfda_logistic_theta1 * ( tw - Qfda_logistic_theta2 ) ) ) ;

Qfartb = Qflvtl + Qfda ;

Qfcardiac = Qfartb ;

Qflung = Qfrvtl - Qfda ;

Qfplacenta = 60 * 24 * 0.001 * Qfplacenta_logistic_theta0 / ( 1 + exp ( - Qfplacenta_logistic_theta1 * ( tw - Qfplacenta_logistic_theta2 ) ) ) ;

Qfdv = 60 * 24 * 0.001 * Qfdv_gompertz_theta0 * exp ( Qfdv_gompertz_theta1 / Qfdv_gompertz_theta2 * ( 1 - exp ( - Qfdv_gompertz_theta2 * tw ) ) ) ;

Qfgut = Qfgut_percent / Qfnonplacental_percent * ( 1 - Qfplacenta / Qfartb ) * Qfartb ;

Qfkidney = Qfkidney_percent / Qfnonplacental_percent * ( 1 - Qfplacenta / Qfartb ) * Qfartb ;

Qfbrain = Qfbrain_percent / Qfnonplacental_percent * ( 1 - Qfplacenta / Qfartb ) * Qfartb ;

Qfliver = Qfliver_percent / ( 100 - ( Qbrain_percent + Qkidney_percent + Qgut_percent ) ) * ( 1 - ( Qfbrain_percent +Qfkidney_percent + Qfgut_percent ) / Qfnonplacental_percent ) * ( 1 - Qfplacenta / Qfartb ) * Qfartb ;

Qfthyroid = Qfthyroid_percent / ( 100 - ( Qbrain_percent + Qkidney_percent + Qgut_percent ) ) * ( 1 - ( Qfbrain_percent +Qfkidney_percent + Qfgut_percent ) / Qfnonplacental_percent ) * ( 1 - Qfplacenta / Qfartb ) * Qfartb ;

Qfrest = Qfcardiac - ( Qfplacenta + Qfgut + Qfliver + Qfthyroid + Qfkidney + Qfbrain ) ;

Qfbypass = Qfcardiac - Qflung ;

Value

list containing:

Author(s)

John Wambaugh

References

Kapraun DF, Wambaugh JF, Setzer RW, Judson RS (2019).“Empirical models for anatomical and physiological changes in a human mother and fetus during pregnancy and gestation.”PLOS ONE,14(5), 1-56.doi:10.1371/journal.pone.0215906.

Calculate the correction for lipid binding in plasma binding assay

Description

Poulin and Haddad (2012) observed "...that fora highly lipophilic compound, the calculated f_up is byfar [less than] the experimental values observed underin vitro conditions." Pearce et al. (2017) hypothesized that there was additional lipidbinding in vivo that acted as a sink for lipophilic compounds, reducing theeffective f_up in vivo. It is possible that this is due to the binding of lipophilic compounds on the nonplasma-side of the rapid equilibrium dialysis plates (Waters et al., 2008).Pearce et al. (2017) compared predicted and observed tissue partition coefficientsfor a range of compounds. They showed that predictions were improved by adding additional binding proportional to the distribution coefficient D_ow (calc_dow)and the fractional volume of lipid inplasma (F_lipid). We calculateF_lipid as the sum of the physiological plasma neutral lipid fractional volume and 30 percent of the plasma neutral phospholipid fractional volume. We use valuesfrom Peyret et al. (2010) for rats and Poulin and Haddad (2012)for humans. The estimate of 30 percent of theneutral phospholipid volume as neutral lipid was used for simplictity's sake inplace of our membrane affinity predictor. To account for additional binding to lipid, plasma to water partitioning(K_plasma:water = 1/f_up)is increased as such:K^corrected_plasma:water = 1/f^corrected_up = 1/f^{in vitro}_up + D_ow*F_lipid

Usage

calc_fup_correction(  fup = NULL,  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  Flipid = NULL,  plasma.pH = 7.4,  dow74 = NULL,  species = "Human",  default.to.human = FALSE,  force.human.fup = FALSE,  suppress.messages = FALSE)

Arguments

Details

Note that octanal:water partitioning above 1:1,000,000 (LogD_ow > 6)are truncated at 1:1,000,000 because greater partitioning wouldlikely take longer than protein binding assay itself.

Value

A numeric fraction unpbound in plasma between zero and one

Author(s)

John Wambaugh

References

Pearce RG, Setzer RW, Davis JL, Wambaugh JF (2017).“Evaluation and calibration of high-throughput predictions of chemical distribution to tissues.”Journal of pharmacokinetics and pharmacodynamics,44, 549–565.doi:10.1007/s10928-017-9548-7.

Peyret T, Poulin P, Krishnan K (2010).“A unified algorithm for predicting partition coefficients for PBPK modeling of drugs and environmental chemicals.”Toxicology and applied pharmacology,249(3), 197–207.doi:10.1016/j.taap.2010.09.010.

Poulin P, Haddad S (2012).“Advancing prediction of tissue distribution and volume of distribution of highly lipophilic compounds from a simplified tissue-composition-based model as a mechanistic animal alternative method.”Journal of pharmaceutical sciences,101(6), 2250–2261.doi:10.1002/jps.23090.

Schmitt W (2008).“General approach for the calculation of tissue to plasma partition coefficients.”Toxicology in vitro,22(2), 457–467.doi:10.1016/j.tiv.2007.09.010.

Waters NJ, Jones R, Williams G, Sohal B (2008).“Validation of a rapid equilibrium dialysis approach for the measurement of plasma protein binding.”Journal of pharmaceutical sciences,97(10), 4586–4595.doi:10.1002/jps.21317.

See Also

apply_fup_adjustment

calc_dow

Calculates the half-life for a one compartment model.

Description

This function calculates the half life from the three compartmentsteady state model where elimination is entirely due to metabolism by theliver and glomerular filtration in the kidneys.

Usage

calc_half_life(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  species = "Human",  model = "3compartmentss",  suppress.messages = TRUE,  ...)

Arguments

Details

Half life is calculated by dividing the natural-log of 2 by the eliminationrate from the one compartment model.

Value

Author(s)

Sarah E. Davidson

See Also

calc_elimination_rate

Examples

calc_half_life(chem.name="Bisphenol A")calc_half_life(chem.name="Bisphenol A",species="Rat")calc_half_life(chem.cas="80-05-7")# Volatiles are outside the domain of default model:try(calc_half_life(     chem.name="toluene"))# We can turn off physchem checking:calc_half_life(     chem.name="toluene",     physchem.exclude=FALSE)# Or use an appropriate model for volatiles:calc_half_life(     chem.name="toluene",     model="sumclearances")# PFAS are outside the domain:try(calc_half_life(     dtxsid="DTXSID8031865",     model="sumclearances"))# Can turn off chemical class checking:calc_half_life(  dtxsid="DTXSID8031865",  model="sumclearances",  class.exclude=FALSE,  suppress.messages=TRUE)# For a metabolized compound, non-restrictive clearance should be faster:h1 <- calc_half_life(  chem.name="toluene",  model="sumclearances",  suppress.messages=TRUE)h2 <- calc_half_life(  chem.name="toluene",  model="sumclearances",  restrictive.clearance=FALSE,  suppress.messages=TRUE)# Check that h2 < h1:if (!(h2 < h1)) stop("h2 not less than h1")# Change species: calc_half_life(  dtxsid="DTXSID8031865",  species="rat",  model="sumclearances",  default.to.human=TRUE,  class.exclude=FALSE,  physchem.exclude=FALSE,  suppress.messages=TRUE)

Calculate first pass heaptic metabolism

Description

For models that don't described first pass blood flow from the gut, need tocacluate a hepatic bioavailability, that is, the fraction of chemical systemically available after metabolism during the first pass through the liver (Rowland, 1973 Equation 29, where k21 is blood flow through the liverand k23 is clearance from the liver in Figure 1 in that paper).

Usage

calc_hep_bioavailability(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  restrictive.clearance = TRUE,  default.to.human = FALSE,  flow.34 = TRUE,  suppress.messages = FALSE,  species = "Human")

Arguments

Value

A data.table whose columns are the parameters of the HTTK modelspecified inmodel.

Author(s)

John Wambaugh

References

Rowland M, Benet LZ, Graham GG (1973).“Clearance concepts in pharmacokinetics.”Journal of pharmacokinetics and biopharmaceutics,1(2), 123–136.doi:10.1007/BF01059626.

Calculate the hepatic clearance.

Description

This function calculates the hepatic clearance in plasma for using the"well-stirred" model by default. Other scaling options from (Ito and Houston 2004) are also available.Parameters for scaling from flow-free intrinsic-hepatic clearance to whole-liver metabolism rate are taken from (Carlile et al. 1997).In vitro measured hepatic clearace is corrected for estimated binding in the in vitro clearance assay using the model of (Kilford et al. 2008).The agument restrictive.clearance (defaults to TRUE) describes thesignificance (or lack thereof) of plasma protein binding in metabolism. Restrictive clearance assumes that only the free fraction of chemical inplasma is available for metabolism. Non-restrictive clearance assumes that the compound is weakly bound toplasma protein and any free chemical metabolized is instantlyreplaced. For non-restrictive clearance the effective fup = 1.

Usage

calc_hep_clearance(  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  parameters = NULL,  hepatic.model = "well-stirred",  suppress.messages = FALSE,  well.stirred.correction = TRUE,  restrictive.clearance = TRUE,  species = "Human",  adjusted.Funbound.plasma = TRUE,  ...)

Arguments

Value

Author(s)

John Wambaugh and Robert Pearce

References

Carlile DJ, Zomorodi K, Houston JB (1997).“Scaling factors to relate drug metabolic clearance in hepatic microsomes, isolated hepatocytes, and the intact liver: studies with induced livers involving diazepam.”Drug metabolism and disposition,25(8), 903–911.

Ito K, Houston JB (2004).“Comparison of the use of liver models for predicting drug clearance using in vitro kinetic data from hepatic microsomes and isolated hepatocytes.”Pharmaceutical research,21, 785–792.doi:10.1023/B:PHAM.0000026429.12114.7d.

Kilford PJ, Gertz M, Houston JB, Galetin A (2008).“Hepatocellular binding of drugs: correction for unbound fraction in hepatocyte incubations using microsomal binding or drug lipophilicity data.”Drug Metabolism and Disposition,36(7), 1194–1197.doi:10.1124/dmd.108.020834.

Pearce RG, Setzer RW, Davis JL, Wambaugh JF (2017).“Evaluation and calibration of high-throughput predictions of chemical distribution to tissues.”Journal of pharmacokinetics and pharmacodynamics,44, 549–565.doi:10.1007/s10928-017-9548-7.

Yang J, Jamei M, Yeo KR, Rostami-Hodjegan A, Tucker GT (2007).“Misuse of the well-stirred model of hepatic drug clearance.”Drug Metabolism and Disposition,35(3), 501–502.doi:10.1124/dmd.106.013359.

Examples

calc_hep_clearance(chem.name="Ibuprofen",hepatic.model='unscaled')calc_hep_clearance(chem.name="Ibuprofen",well.stirred.correction=FALSE)

Calculate the free chemical in the hepaitic clearance assay

Description

This function uses the method from Kilford et al. (2008) to calculate thefraction of unbound chemical in the hepatocyte intrinsic clearance assay. The bound chemical is presumed to beunavailable during the performance of the assay, so this fraction can beused to increase the apparent clearance rate to better estimate in vivo clearance. For bases, the fraction of chemical unbound in hepatocyte clearance assays (fu_hep) is calculated in terms of logP_owbut for neutrual and acidic compounds we use logD_ow (fromcalc_dow). Here we denote the appropriate partition coefficient as "logP/D".Kilford et al. (2008) calculatesfu_hep = 1/(1 + 125*V_R*10^(0.072*logP/D² + 0.067*logP/D-1.126))

Usage

calc_hep_fu(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  Vr = 0.005,  pH = 7.4)

Arguments

Details

Note that octanal:water partitioning above 1:1,000,000 (LogP_ow > 6)are truncated at 1:1,000,000 because greater partitioning wouldlikely take longer than hepatocyte assay itself.

Value

A numeric fraction between zero and one

Author(s)

John Wambaugh and Robert Pearce

References

Kilford PJ, Gertz M, Houston JB, Galetin A (2008).“Hepatocellular binding of drugs: correction for unbound fraction in hepatocyte incubations using microsomal binding or drug lipophilicity data.”Drug Metabolism and Disposition,36(7), 1194–1197.doi:10.1124/dmd.108.020834.

Wetmore BA, Wambaugh JF, Allen B, Ferguson SS, Sochaski MA, Setzer RW, Houck KA, Strope CL, Cantwell K, Judson RS, others (2015).“Incorporating high-throughput exposure predictions with dosimetry-adjusted in vitro bioactivity to inform chemical toxicity testing.”Toxicological Sciences,148(1), 121–136.doi:10.1093/toxsci/kfv171.

See Also

apply_clint_adjustment

Calculate the hepatic clearance (deprecated).

Description

This function is included for backward compatibility. It callscalc_hep_clearance whichcalculates the hepatic clearance in plasma for a well-stirred modelor other type if specified. Based on Ito and Houston (2004)

Usage

calc_hepatic_clearance(  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  parameters = NULL,  species = "Human",  default.to.human = FALSE,  hepatic.model = "well-stirred",  suppress.messages = FALSE,  well.stirred.correction = TRUE,  restrictive.clearance = TRUE,  adjusted.Funbound.plasma = TRUE,  ...)

Arguments

Value

Author(s)

John Wambaugh and Robert Pearce

References

Ito, K., & Houston, J. B. (2004). "Comparison of the use of liver models for predicting drug clearance using in vitro kinetic data from hepatic microsomes and isolated hepatocytes." Pharmaceutical Tesearch, 21(5), 785-792.

Examples

calc_hep_clearance(chem.name="Ibuprofen",hepatic.model='unscaled')calc_hep_clearance(chem.name="Ibuprofen",well.stirred.correction=FALSE)

Calculate the ionization.

Description

This function calculates the ionization of a compound at a given pH. The pKa's are either entered as parameters or taken from a specific compound inthe package. The arguments pKa_Donor and pKa_Accept may be single numbers, characters, or vectors. We support characters because there are many instances with multiple predicted values and all those values can be included by concatenating with commas (for example, pKa_Donor = "8.1,8.6". Finally, pka_Donor and pKa_Accept may be vectors of characters representing different chemicals or instances ofchemical parameters to allow for uncertainty analysis. A null value forpKa_Donor or pKa_Accept is interpretted as no argument provided, while " " is taken as a prediction of no ionization possible at any pH.

Usage

calc_ionization(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  pH = NULL,  pKa_Donor = NULL,  pKa_Accept = NULL,  return_charge_matrix = FALSE)

Arguments

Details

It is very important to note that pKb = 14 - pKa. But if a predictor gives usa doinor pKa, we just accept it as a pKa.

For hydrogen donor sites, a hydrogen is present in the molecule that can be donated to the solution if the concentration of hydrogens gets low enough. This causes the molecule to become more negatively charged. This is an acid. For hydrogen acceptor suits a location exist in the molecule that can accept an additional history if the concentration of hydrogens gets sufficiently high. This causes the molecule to become more positively charged. This is a base.

We make several assumptions about ionization in order to make our calculations.First, we assume ionization is either due to either "donating" (losing) ahydrogen ion (a positively charge proton) to the solution or by "accepting"(gaining) a hydrogen ion from the solution. Generally, acids are hydrogen donorsand bases are hydrogen acceptors. Second, pH is the negative log10 concentrationof hydrogen atoms. The lower the pH, the more hydrogen atoms. So, acids donatetheir hydrogen atoms as pH of the solution increases. Bases accept their hydrogenatoms as pH decreases. Third, each predicted pKa is a prediction that a specificlocation (or site) on molecule X can either donate or accept a hydrogen. Fourth, the pKavalue indicates the pH at which half of the molecules of X have ionized atthe site, and half have not. The concentration of the two forms are equal.Fifth, if there are N pKa's for molecule X, then there are N sites that canionize. Technically this means that there are 2^N different ionization statesfor molecule X (where each site is or is not ionized). However, pKa predictorsgive the equlibrium only for pairs of ionization states. So, we only considerN + 1 ionizations states for X – the state immediately above and below eachpKa.

To understand the different charge states we annotate the nonionizable backboneof a molecule as "X". For each site on X that is capable of donating a hydrogenwe add a "D" to the right of "X". For each site on X that has accepted a hydrogen,we add a "A" to the right of "X". We read the A's and D's from left to right,with the one occuring at the lowest pH first. So a typical acid ionization would be:XD -> X- and a typical base ionization would be XA+ -> X. Where things get complicatedis if there are multiple donor and acceptor states. In particular, it is possible for a compound to have a net zero charge, but be simultaneously positively andnegatively charged. Such a state is called a Zwitter ion. For example:XDAA++ -> XAA+ -> XA -> X-. The state XA is technically neutral because' X has donated one hydrogen, but also accepted one hydrogen. XA is a Zwitter ion.

Each pKa gives the equlibrium ratio of two states pH - pKa = log10[X/XD] fordonation or pOH - pka = log10[X/XA] for accepting. pOH = 14 - pH. Separating thelogarithm into log10[X] - log10[XD] lets us see that Cn = Xn - Xn-1 whereCn = pH -pKa for donor pKa's and Cn = 14 - pH - pKa for acceptor pKa's.We can rewrite log10Xn = Sum_i=1:n Ci + log10X1. So we can calculate each Xnby summing all the ratios between Xn and the lowest state (X1). Then, by requiring that all Xi sum to 1, we have:1 = Sum_i=1:N 10^Xi = Sum_i=1:N 10^(Sum_j=1:i (Cj + log10X1)) = X1 * Sum_i=1:N 10^(Sum_j=1:i Cj)so that X1 = 1 / Sum_i=1:N 10^(Sum_j=1:i Cj)

The sum im the denominator is the ratio from X1 to each state (including X1).We use a table called "charge_matrix" to keep track of all N + 1 ionizationstates and the ratio of each state to the next.

Value

Author(s)

Robert Pearce and John Wambaugh

References

Pearce RG, Setzer RW, Davis JL, Wambaugh JF (2017).“Evaluation and calibration of high-throughput predictions of chemical distribution to tissues.”Journal of pharmacokinetics and pharmacodynamics,44, 549–565.doi:10.1007/s10928-017-9548-7.

Strope CL, Mansouri K, Clewell III HJ, Rabinowitz JR, Stevens C, Wambaugh JF (2018).“High-throughput in-silico prediction of ionization equilibria for pharmacokinetic modeling.”Science of The Total Environment,615, 150–160.doi:10.1016/j.scitotenv.2017.09.033.

Examples

# Neutral compound:calc_ionization(chem.name="Acetochlor",pH=7.4)# Donor pKa's 9.78,10.39 -- Should be almost all neutral at plasma pH:out <- calc_ionization(chem.name='bisphenola',pH=7.4)print(out)out[["fraction_neutral"]]==max(unlist(out))# Donor pKa's 9.78,10.39 -- Should be almost all negative (anion) at higher pH:out <- calc_ionization(chem.name='bisphenola',pH=11)print(out)out[["fraction_negative"]]==max(unlist(out))# Ficticious compound, should be almost all all negative (anion):out <- calc_ionization(pKa_Donor=8,pKa_Accept="1,4",pH=9)print(out)out[["fraction_negative"]]>0.9# Donor pKa 6.54 -- Should be mostly negative (anion):out <- calc_ionization(chem.name='Acephate',pH=7)print(out)out[["fraction_negative"]]==max(unlist(out))#Acceptor pKa's "9.04,6.04"  -- Should be almost all positive (cation) at plasma pH:out <- calc_ionization(chem.cas="145742-28-5",pH=7.4)print(out)out[["fraction_positive"]]==max(unlist(out))#Ficticious Zwitteron:out <- calc_ionization(pKa_Donor=6,pKa_Accept="8",pH=7.4)print(out)out[["fraction_zwitter"]]==max(unlist(out))

Calculate air:matrix partition coefficients

Description

This function uses the methods colleced by Linakis et al. (2020) to calculateair partition coefficients for blood, water, and mucus.

Usage

calc_kair(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  species = "Human",  adjusted.Funbound.plasma = TRUE,  fup.lod.default = 0.005,  force.human.clint.fup = FALSE,  minimum.Funbound.plasma = 1e-04,  default.to.human = FALSE,  suppress.messages = FALSE,  pH = 7.4,  alpha = 0.001)

Arguments

Details

The blood:air partition coefficient (PB:A) was calculated as P_B:A = P_B:A * R_B:P / f_upwhere P_B:A is the blood:air partition, RB:P is the blood:plasma partitionratio, fup is the fraction unbound in the plasma, andP_W:A is the water:air partition coefficient:R * T_body / HLC / Pwhere R is the gas constant (8.314 J/mol/K), T_body is the species-specific body temperature (K) fromphysiology.data,HLC is the Henry's Law Constant (atm*m^3 / mol), and P is conversion factor from atmospheres to Pascals (1 atm = 101325 Pa).

In the isopropanol PBTKmodel published by Clewell et al. (2001) it was noted that certain chemicals are likely to be absorbed into the mucus or otherwisetrapped in the upper respiratory tract (URT). Following Scott (2014), the air:mucus partition coefficient (PA:M) calculated aslog₁₀(1/K_water2air) - (log₁₀(P_ow) - 1) * 0.524where Pow is the octanol/water partition coefficient

Value

A named list containing the blood:air, water:air, and mucus:air partition coefficients

Author(s)

John Wambaugh and Matt Linakis

References

Linakis MW, Sayre RR, Pearce RG, Sfeir MA, Sipes NS, Pangburn HA, Gearhart JM, Wambaugh JF (2020).“Development and evaluation of a high-throughput inhalation model for organic chemicals.”Journal of exposure science & environmental epidemiology,30(5), 866–877.doi:10.1038/s41370-020-0238-y.

Clewell III, Harvey J., et al. "Development of a physiologically based pharmacokinetic model of isopropanol and its metabolite acetone." Toxicological Sciences 63.2 (2001): 160-172.

Scott, John W., et al. "Tuning to odor solubility and sorption pattern in olfactory epithelial responses." Journal of Neuroscience 34.6 (2014): 2025-2036.

See Also

calc_dow

Back-calculates the Red Blood Cell to Unbound Plasma Partition Coefficient

Description

Given an observed ratio of chemical concentration in blood to plasma, thisfunction calculates a Red Blood Cell to unbound plasma (Krbc2pu) partitioncoefficient that would be consistent with that observation.

Usage

calc_krbc2pu(  Rb2p,  Funbound.plasma,  hematocrit = NULL,  default.to.human = FALSE,  species = "Human",  suppress.messages = TRUE)

Arguments

Value

The red blood cell to unbound chemical in plasma partition coefficient.

Author(s)

John Wambaugh and Robert Pearce

References

Pearce RG, Setzer RW, Davis JL, Wambaugh JF (2017).“Evaluation and calibration of high-throughput predictions of chemical distribution to tissues.”Journal of pharmacokinetics and pharmacodynamics,44, 549–565.doi:10.1007/s10928-017-9548-7.

Ruark CD, Hack CE, Robinson PJ, Mahle DA, Gearhart JM (2014).“Predicting passive and active tissue: plasma partition coefficients: interindividual and interspecies variability.”Journal of pharmaceutical sciences,103(7), 2189–2198.doi:10.1002/jps.24011.

Calculate the membrane affinity

Description

Membrane affinity (MA) is the membrane:water partition coefficient. MAchacterizes chemical partitioning into membranes formedfrom neutral phospholipids (K_nPL). Pearce et al. (2017) compared five different methods for predicting membrane affinity using measured data for 59 compounds. The method ofYun and Edgington (2013) was identified as the best: MA = 10^(1.294 + 0.304 * log₁₀(P_ow)

Usage

calc_ma(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  suppress.messages = FALSE,  pfas.calibration = TRUE)

Arguments

Value

A membrane:unbound fraction in plasma partition coefficient

Author(s)

John Wambaugh

References

Pearce RG, Setzer RW, Davis JL, Wambaugh JF (2017).“Evaluation and calibration of high-throughput predictions of chemical distribution to tissues.”Journal of pharmacokinetics and pharmacodynamics,44, 549–565.doi:10.1007/s10928-017-9548-7.

Yun YE, Edginton AN (2013).“Correlation-based prediction of tissue-to-plasma partition coefficients using readily available input parameters.”Xenobiotica,43(10), 839–852.doi:10.3109/00498254.2013.770182.

Droge STJ (2019).“Membrane?Water Partition Coefficients to Aid Risk Assessment of Perfluoroalkyl Anions and Alkyl Sulfates.”Environmental Science & Technology,53(2), 760-770.doi:10.1021/acs.est.8b05052.PMID: 30572703, https://doi.org/10.1021/acs.est.8b05052.

Calculate maternal body weight

Description

This function initializes the parameters needed in the functionssolve_fetal_pbtk by calling solve_pbtk and adding additional parameters.

Usage

calc_maternal_bw(week = 12)

Arguments

Details

BW <- params$pre_pregnant_BW + params$BW_cubic_theta1 * tw + params$BW_cubic_theta2 * tw^2 + params$BW_cubic_theta3 * tw^3

Value

Author(s)

John Wambaugh

References

Kapraun DF, Wambaugh JF, Setzer RW, Judson RS (2019).“Empirical models for anatomical and physiological changes in a human mother and fetus during pregnancy and gestation.”PLOS ONE,14(5), 1-56.doi:10.1371/journal.pone.0215906.

Distribution of chemical steady state concentration with uncertainty and variability

Description

For a given chemical and fixed dose rate this function determines a distribution of steady-state concentrations reflecting measurement uncertaintyan population variability. Uncertainty and variability are simulated via theMonte Carlo method – many sets of model parameters are drawn according toprobability distributions described in Ring et al. (2017) (doi:10.1016/j.envint.2017.06.004)for human variability and Wambaugh et al. (2019) (doi:10.1093/toxsci/kfz205) for measurement uncertainty. Monte Carlo samples are generated by the functioncreate_mc_samples. To allow rapid application of the Monte Carlo method we make use of analytical solutions for the steady-state concentrationfor a particular model via a given route (when available) as opposed tosolving the model numerically (that is, using differential equations).For each sample of the Monte Carlo method (as specified by argumentsamples) the parameters for the analytical solution are varied. Anensemble of steady-state predictions are produced, though by default onlythe quantiles specified by argumentwhich.quantile are provided. If the full set of predicted values are desired use set the argumentreturn.samples toTRUE.

Usage

calc_mc_css(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  samples = 1000,  which.quantile = 0.95,  species = "Human",  daily.dose = 1,  suppress.messages = FALSE,  model = "3compartmentss",  httkpop = TRUE,  httkpop.dt = NULL,  invitrouv = TRUE,  calcrb2p = TRUE,  censored.params = list(),  vary.params = list(),  return.samples = FALSE,  tissue = NULL,  concentration = "plasma",  output.units = "mg/L",  invitro.mc.arg.list = NULL,  httkpop.generate.arg.list = list(method = "direct resampling"),  convert.httkpop.arg.list = NULL,  parameterize.args.list = NULL,  calc.analytic.css.arg.list = NULL,  Caco2.options = NULL)

Arguments

Details

The chemical-specific steady-state concentration for a dose rate of 1 mg/kg body weight/day can be used for inin vitro-in vivo extrapolation (IVIVE) ofa bioactivein vitro concentration by dividing thein vitroconcentrationby the steady-state concentration to predict a dose rate (mg/kg/day) thatwould produce that concentration in plasma. Using quantiles of the distribution (such as the upper 95th percentile) allow incorporation of uncertainty and variability into IVIVE.

Reverse Dosimetry Toxicodynamic IVIVE

Figure from Breen et al. (2021) (doi:10.1080/17425255.2021.1935867) shows reverse dosimetry toxicodynamic IVIVE. Equivalent external dose isdetermined by solving the TK model in reverse by deriving the external dose(that is, TK model input) that produces a specified internal concentration (that is, TK model output). Reverse dosimetry and IVIVE using HTTK relies on the linearity of the models. We calculate a scaling factor to relatein vitro concentrations (uM) to administered equivalent doses (AED). The scaling factor is the inverse of the steady state plasma concentration (Css) predicted for a 1 mg/kg/day exposure dose rate. We use Monte Carlo to simulate variability and propagate uncertainty; for example, to calculate an upper 95th percentile Css,95 for individuals who get higher plasma concentrations from the same exposure.

The Monte Carlo methods used here were recently updated and described byBreen et al. (submitted).

httk-pop is used only for humans. For non-human species biological variability is simulated by drawing parameters from uncorellated log-normaldistributions.

Chemical-specific httk data are available primarily for human and for a fewhundred chemicals in rat. All in silico predictions are for human. Thus, when species is specified as rabbit, dog, or mouse, the user can chooseto set the argumentdefault.to.human toTRUE so that this function uses theappropriate physiological data (volumes and flows) but substitutes humanfraction unbound, partition coefficients, and intrinsic hepatic clearance.

If the argumenttissue is used, the steady-state concentration in thattissue, if available, is provided. If that tissue is included in the modelused (specified by arguementmodel) then the actual tissue concentration isprovided. Otherwise, the tissue-specific partition coefficient is used toestimate the concentration from the plasma.

The six sets of plausible IVIVEassumptions identified by Honda et al. (2019) (doi:10.1371/journal.pone.0217564) are:

*Assumption iscurrently ignored because analytical steady-state solutions are currentlyused by this function.

Value

Quantiles (specified by which.quantile) of the distribution of plasmasteady-stae concentration (Css) from the Monte Carlo simulation

Author(s)

Caroline Ring, Robert Pearce, John Wambaugh, Miyuki Breen, and Greg Honda

References

Wambaugh JF, Wetmore BA, Pearce R, Strope C, Goldsmith R, Sluka JP, Sedykh A, Tropsha A, Bosgra S, Shah I, others (2015).“Toxicokinetic triage for environmental chemicals.”Toxicological Sciences,147(1), 55–67.doi:10.1093/toxsci/kfv118.

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

Honda GS, Pearce RG, Pham LL, Setzer RW, Wetmore BA, Sipes NS, Gilbert J, Franz B, Thomas RS, Wambaugh JF (2019).“Using the concordance of in vitro and in vivo data to evaluate extrapolation assumptions.”PloS one,14(5), e0217564.doi:10.1371/journal.pone.0217564.

Rowland M, Benet LZ, Graham GG (1973).“Clearance concepts in pharmacokinetics.”Journal of pharmacokinetics and biopharmaceutics,1(2), 123–136.doi:10.1007/BF01059626.

Breen M, Wambaugh JF, Bernstein A, Sfeir M, Ring CL (2022).“Simulating toxicokinetic variability to identify susceptible and highly exposed populations.”Journal of Exposure Science & Environmental Epidemiology,32(6), 855–863.doi:10.1038/s41370-022-00491-0.

See Also

calc_analytic_css

create_mc_samples

Examples

# Set the number of samples (NSAMP) to a small value for rapid testing, # increase NSAMP for more stable (reproducible) results. Default value is 1000:NSAMP = 100# Basic in vitro - in vivo extrapolation with httk, convert 3 uM in vitro# concentration of chemical with CAS 2451-62-9 to mg/kg/day:set.seed(1234)3/calc_mc_css(chem.cas="2451-62-9", samples=NSAMP, output.units="uM")# The significant digits should give the same answer as:set.seed(1234)calc_mc_oral_equiv(chem.cas="2451-62-9", conc=3, samples=NSAMP)  # By default human variability is simulated using httk-pop based upon actual# individuals from three recent NHANES cohorts (see Breen et al., 2022):set.seed(1234)calc_mc_css(chem.name='Bisphenol A', output.units='uM',            samples=NSAMP, return.samples=TRUE)# However, as described in Ring et al. (2017) we can also sample using# virtual individuals drawn from distributions determiend from the NHANES# cohorts (this is method "vi"):set.seed(1234)calc_mc_css(chem.name='Bisphenol A', output.units='uM',            samples=NSAMP,            httkpop.generate.arg.list=list(method='vi'))                          # We can also tailor the NHANES cohort to provide variability simulation of # specific populations:set.seed(1234)calc_mc_css(chem.cas = "80-05-7",             which.quantile = 0.5,            output.units = "uM",             samples = NSAMP,            httkpop.generate.arg.list=list(method='vi',                                            gendernum=NULL,                                            agelim_years=NULL,                                           agelim_months=NULL,                                           weight_category = c("Underweight",                                                               "Normal",                                                               "Overweight",                                                               "Obese"                                                               )                                           )            )# The following example should result in an error since we do not # estimate tissue partitioning with '3compartmentss', which does not have# tissues:                         set.seed(1234)try(calc_mc_css(chem.name='2,4-d',                 which.quantile=.9,                samples=NSAMP,                httkpop=FALSE,                 tissue='heart')) # But heart will work with PBTK, even though it's lumped since we estimate# a partition coefficient before lumping:set.seed(1234)calc_mc_css(chem.name='2,4-d', model='pbtk',            samples=NSAMP,            which.quantile=.9, httkpop=FALSE, tissue='heart')# It is also possible to use this function with a vector of model parameters,# although you must specify the model to which the parameters correspond:params <- parameterize_pbtk(chem.cas="80-05-7")set.seed(1234)calc_mc_css(parameters=params,model="pbtk", samples=NSAMP)# By default the standard HTTK Monte Carlo uses httk-pop (Ring et al., 2017):set.seed(1234)calc_mc_css(chem.cas="90-43-7", model="pbtk", samples=NSAMP)# We can use HTTK Monte Carlo with no measurement uncertainty (this is how # monte carlo was performed before v1.10.0):set.seed(1234)calc_mc_css(chem.cas="90-43-7",            model="pbtk",            samples=NSAMP,            invitro.mc.arg.list = list(            adjusted.Funbound.plasma = TRUE,            poormetab = TRUE,             fup.censored.dist = FALSE,             fup.lod = 0.01,             fup.meas.cv = 0.0,             clint.meas.cv = 0.0,             fup.pop.cv = 0.3,             clint.pop.cv = 0.3)) # HTTK Monte Carlo with no HTTK-Pop physiological variability):set.seed(1234)calc_mc_css(chem.cas="90-43-7",model="pbtk",samples=NSAMP,httkpop=FALSE)# HTTK Monte Carlo with no in vitro uncertainty and variability):set.seed(1234)calc_mc_css(chem.cas="90-43-7",model="pbtk",samples=NSAMP,invitrouv=FALSE)# HTTK Monte Carlo with no HTTK-Pop and no in vitro uncertainty and variability):set.seed(1234)calc_mc_css(chem.cas="90-43-7" ,model="pbtk",            samples=NSAMP, httkpop=FALSE, invitrouv=FALSE)# Should be the same as the mean result:calc_analytic_css(chem.cas="90-43-7",model="pbtk",output.units="mg/L")# HTTK Monte Carlo using basic Monte Carlo sampler:set.seed(1234)calc_mc_css(chem.cas="90-43-7",            model="pbtk",            samples=NSAMP,            httkpop=FALSE,            invitrouv=FALSE,            vary.params=list(Pow=0.3)) # The following will not work because Diquat dibromide monohydrate's # Henry's Law Constant (-3.912) is higher than that of Acetone (~-4.5):try(calc_mc_css(chem.cas="6385-62-2"))# However, we can turn off checking for phys-chem properties, since we know# that  Diquat dibromide monohydrate is not too volatile:calc_mc_css(chem.cas="6385-62-2", parameterize.args.list =list(physchem.exclude=FALSE))# We can also use the Monte Carlo functions by passing a table# where each row represents a different Monte Carlo draw of parameters:p <- create_mc_samples(chem.cas="80-05-7")# Use data.table for steady-state plasma concentration (Css) Monte Carlo:calc_mc_css(parameters=p)# Using the same table gives the same answer:calc_mc_css(parameters=p)# Use Css for 1 mg/kg/day for simple reverse toxicokinetics # in Vitro-In Vivo Extrapolation to convert 15 uM to mg/kg/day:15/calc_mc_css(parameters=p, output.units="uM")# Can do the same with calc_mc_oral_equiv:calc_mc_oral_equiv(15, parameters=p)# HTTK Monte Carlo using basic Monte Carlo sampler:set.seed(1234)calc_mc_css(chem.cas="90-43-7",  model="pbtk",  samples=NSAMP,  httkpop=FALSE,  invitrouv=FALSE,  vary.params=list(Pow=0.3))# make sure the oral equivalent function works:set.seed(1234)calc_mc_oral_equiv(chem.name="bisphenol a",conc=10,samples=NSAMP)set.seed(1234)# Do the calculation manually to make sure units are correct:signif(10/calc_mc_css(chem.name="bisphenol a",samples=NSAMP,output.units="uM"),4)# do test of passing data.table of parametersset.seed(1234)parameter.dt <- create_mc_samples(chem.cas="335104-84-2",                                    model="pbtk",                                    samples=NSAMP)calc_mc_oral_equiv(conc=100,                   parameters=parameter.dt,                   model="pbtk",                   samples=NSAMP)# do test of passing single set of parametersparams <- parameterize_steadystate(chem.cas="80-05-7")css3 <- calc_analytic_css(  parameters=params,  output.units = "uM",   model = "3compartmentss",   species = "Human")set.seed(1234)css4 <- calc_mc_css(  parameters=params,   output.units = "uM",   model = "3compartmentss",   species = "Human",   httkpop=FALSE,   invitrouv=FALSE,   return.samples=TRUE,  samples=NSAMP)set.seed(1234)css5 <- calc_mc_css(  parameters=params,   output.units = "uM",   model = "3compartmentss",   species = "Human",   httkpop=TRUE,   invitrouv=TRUE,   return.samples=TRUE,  samples=NSAMP)# If we turn off all the montecarlo the samples should all be the same and# give us the same result as calc_analytic_css:set.seed(1234)css1 <- calc_mc_css(  chem.cas = "80-05-7",   output.units = "uM",   model = "3compartmentss",   species = "Human",   httkpop=FALSE,   invitrouv=FALSE,   return.samples=TRUE,  samples=NSAMP)set.seed(1234)css2 <- calc_analytic_css(  chem.cas = "80-05-7",   output.units = "uM",   model = "3compartmentss",   species = "Human")# These values should be the same:all(mean(abs(signif((css1-css2)/css2,4)))<0.001)css2/css3==1# Because we can't recalculate Rblood2plasma for 3compartmentss this is not# quite the same but should be close:unique(css4)/css2# Now test that MC works across different models:set.seed(1234)calc_mc_css(chem.cas="15972-60-8",model="3compartment",samples=NSAMP)set.seed(1234)calc_mc_css(chem.cas="15972-60-8",model="1compartment",samples=NSAMP)set.seed(1234)calc_mc_css(chem.cas="15972-60-8",model="pbtk",samples=NSAMP)# Should be the same as the mean result:calc_analytic_css(chem.cas="90-43-7",model="pbtk",output.units="mg/L")

Calculate Monte Carlo Oral Equivalent Dose

Description

This function converts a chemical plasma concentration to an oral administeredequivalentdose (AED) using a concentration obtained fromcalc_mc_css. This function uses reverse dosimetry-based 'in vitro-in vivo extrapolation (IVIVE) for high throughput risk screening. The user can input thechemical andin vitro bioactive concentration, select the TKmodel, and then automatically predict thein vivo AEDwhich would produce a body concentration equal tothein vitro bioactive concentration. This function relies on theMonte Carlo method (via functioncreate_mc_samples to simulateboth uncertainty and variability so that the result is a distribution ofequivalent doses, from which we provide specific quantiles(specified bywhich.quantile), though the full set of predictions canbe obtained by settingreturn.samples toTRUE.

Usage

calc_mc_oral_equiv(  conc,  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  parameters = NULL,  which.quantile = 0.95,  species = "Human",  input.units = "uM",  output.units = "mgpkgpday",  suppress.messages = FALSE,  return.samples = FALSE,  restrictive.clearance = TRUE,  bioactive.free.invivo = FALSE,  tissue = NULL,  concentration = "plasma",  IVIVE = NULL,  model = "3compartmentss",  Caco2.options = list(),  calc.analytic.css.arg.list = list(),  ...)

Arguments

Details

The chemical-specific steady-state concentration for a dose rate of 1 mg/kg body weight/day can be used for in IVIVE ofa bioactivein vitro concentration by dividing thein vitroconcentrationby the steady-state concentration to predict a dose rate (mg/kg/day) thatwould produce that concentration in plasma. Using quantiles of the distribution (such as the upper 95th percentile) allow incorporation of uncertainty and variability into IVIVE.

This approach relies on the linearityof the models to calculate a scaling factor to relatein vitroconcentrations (uM) with AED. The scaling factor is theinverse of the steady-state plasma concentration (Css) predictedfor a 1 mg/kg/day exposure dose ratewherein vitro concentration [X] and Css must be in the sameunits. Note that it is typical forin vitro concentrations to bereported in units of uM and Css in units of mg/L, in which caseone must be converted to the other.

Reverse Dosimetry Toxicodynamic IVIVE

Figure from Breen et al. (2021) (doi:10.1080/17425255.2021.1935867) shows reverse dosimetry toxicodynamic IVIVE. Equivalent external dose isdetermined by solving the TK model in reverse by deriving the external dose(that is, TK model input) that produces a specified internal concentration (that is, TK model output). Reverse dosimetry and IVIVE using HTTK relies on the linearity of the models. We calculate a scaling factor to relatein vitro concentrations (uM) to AEDs. The scaling factor is the inverse of the Css predicted for a 1 mg/kg/day exposure dose rate. We use Monte Carlo to simulate variability and propagate uncertainty; for example, to calculate an upper 95th percentile Css,95 for individuals who get higher plasma concentrations from the same exposure.

The Monte Carlo methods used here were recently updated and described byBreen et al. (submitted).

All arguments after httkpop only apply if httkpop is set to TRUE and speciesto "Human".

When species is specified as rabbit, dog, or mouse, the function uses theappropriate physiological data(volumes and flows) but substitutes humanfraction unbound, partition coefficients, and intrinsic hepatic clearance.

Tissue concentrations are calculated for the pbtk model with oral infusiondosing. All tissues other than gut, liver, and lung are the product of thesteady state plasma concentration and the tissue to plasma partitioncoefficient.

The six sets of plausible IVIVEassumptions identified by Honda et al. (2019) (doi:10.1371/journal.pone.0217564) are:

*Assumption iscurrently ignored because analytical steady-state solutions are currentlyused by this function.

Value

Equivalent dose in specified units, default of mg/kg BW/day.

Author(s)

John Wambaugh

References

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

Wetmore BA, Wambaugh JF, Allen B, Ferguson SS, Sochaski MA, Setzer RW, Houck KA, Strope CL, Cantwell K, Judson RS, others (2015).“Incorporating high-throughput exposure predictions with dosimetry-adjusted in vitro bioactivity to inform chemical toxicity testing.”Toxicological Sciences,148(1), 121–136.doi:10.1093/toxsci/kfv171.

Honda GS, Pearce RG, Pham LL, Setzer RW, Wetmore BA, Sipes NS, Gilbert J, Franz B, Thomas RS, Wambaugh JF (2019).“Using the concordance of in vitro and in vivo data to evaluate extrapolation assumptions.”PloS one,14(5), e0217564.doi:10.1371/journal.pone.0217564.

Rowland M, Benet LZ, Graham GG (1973).“Clearance concepts in pharmacokinetics.”Journal of pharmacokinetics and biopharmaceutics,1(2), 123–136.doi:10.1007/BF01059626.

See Also

calc_mc_css

create_mc_samples

Examples

# Set the number of samples (NSAMP) low for rapid testing, increase NSAMP # for more stable results. Default value is 1000:NSAMP = 10# Basic in vitro - in vivo extrapolation with httk, convert 0.5 uM in vitro# concentration of chemical Surinabant to mg/kg/day:set.seed(1234)0.5/calc_mc_css(chem.name="Surinabant", samples=NSAMP, output.units="uM")# The significant digits should give the same answer as:set.seed(1234)calc_mc_oral_equiv(chem.name="Surinabant",conc=0.5,samples=NSAMP)  # Note that we use set.seed to get the same sequence of random numbers for# the two different function calls (calc_mc_css and calc_mc_oral_equiv)# The following example should result in an error since we do not # estimate tissue partitioning with '3compartmentss'. set.seed(1234)                        try(calc_mc_oral_equiv(0.1, chem.cas="34256-82-1",                       which.quantile=c(0.05,0.5,0.95),                       samples=NSAMP,                       tissue='brain'))       set.seed(1234)calc_mc_oral_equiv(0.1,chem.cas="34256-82-1", model='pbtk',                   samples=NSAMP,                   which.quantile=c(0.05,0.5,0.95), tissue='brain') # The following will not work because Diquat dibromide monohydrate's # Henry's Law Constant (-3.912) is higher than that of Acetone (~-4.5):try(calc_mc_oral_equiv(3, chem.cas="6385-62-2"))# However, we can turn off checking for phys-chem properties, since we know# that  Diquat dibromide monohydrate is not too volatile:calc_mc_oral_equiv(3, chem.cas="6385-62-2", parameterize.args.list =list(physchem.exclude=FALSE)) # We can also use the Monte Carlo functions by passing a table# where each row represents a different Monte Carlo draw of parameters:p <- create_mc_samples(chem.cas="80-05-7")# Use data.table for steady-state plasma concentration (Css) Monte Carlo:calc_mc_css(parameters=p)# Using the same table gives the same answer:calc_mc_css(parameters=p)# Use Css for 1 mg/kg/day for simple reverse toxicokinetics # in Vitro-In Vivo Extrapolation to convert 15 uM to mg/kg/day:15/calc_mc_css(parameters=p, output.units="uM")# Can do the same with calc_mc_oral_equiv:calc_mc_oral_equiv(15, parameters=p)

Conduct multiple TK simulations using Monte Carlo

Description

Solves a model for concentration vs. time predictions usingMonte Carlo methods

Usage

calc_mc_tk(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  samples = 1000,  species = "Human",  suppress.messages = FALSE,  model = "pbtk",  httkpop = TRUE,  httkpop.dt = NULL,  invitrouv = TRUE,  calcrb2p = TRUE,  censored.params = list(),  vary.params = list(),  return.samples = FALSE,  tissue = NULL,  output.units = "mg/L",  solvemodel.arg.list = list(times = c(0, 0.25, 0.5, 0.75, 1, 1.5, 2, 2.5, 3, 4, 5)),  Caco2.options = list(),  invitro.mc.arg.list = NULL,  httkpop.generate.arg.list = list(method = "direct resampling"),  convert.httkpop.arg.list = NULL,  parameterize.args.list = NULL,  propagate.invitrouv.arg.list = NULL,  return.all.sims = FALSE)

Arguments

Details

The Monte Carlo methods used here were recently updated and described byBreen et al. (submitted).

All arguments after httkpop only apply if httkpop is set to TRUE and speciesto "Human".

When species is specified as rabbit, dog, or mouse, the function uses theappropriate physiological data(volumes and flows) but substitutes humanfraction unbound, partition coefficients, and intrinsic hepatic clearance.

The six sets of plausiblein vitro-in vivo extrapolation (IVIVE)assumptions identified by Honda et al. (2019)(doi:10.1371/journal.pone.0217564)are:

NOTE: This function, and [create_mc_samples()], calculate oralbioavailability parameters based on Caco2 data. If you are comparing theoutput of this function to the output of [solve_model()], you will need to toensure that [solve_model()] also uses Caco2 data, rather than its defaultbehavior of using *in vivo* oral bioavailability data when available. To dothat, please pass the following argument to [solve_model()]: 'Caco2.options =list(overwrite.invivo = TRUE'.

Value

If return.all.sims == FALSE (default) a list with:

If return.all.sums == TRUE then a list is returned with:

Author(s)

John Wambaugh

See Also

create_mc_samples

Examples

NSAMP <- 50chemname="Abamectin"times<- c(0,0.25,0.5,0.75,1,1.5,2,2.5,3,4,5)age.ranges <- seq(6,80,by=10)forward <- NULLfor (age.lower in age.ranges){  label <- paste("Ages ",age.lower,"-",age.lower+4,sep="")  set.seed(1234)  forward[[label]] <- calc_mc_tk(                        chem.name=chemname,                        samples=NSAMP,                        httkpop.generate.arg.list=list(                          method="d",                          agelim_years = c(age.lower, age.lower+9)),                        solvemodel.arg.list = list(                          times=times))}set.seed(1234)# well-behaved chemical with a measured Rblood2plasma:lapply(calc_mc_tk(chem.cas="80-05-7",samples=NSAMP),function(x) x[-2,])

Calculate the constant ratio of the blood concentration to the plasmaconcentration.

Description

This function calculates the constant ratio of the blood concentration tothe plasma concentration.

Usage

calc_rblood2plasma(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  hematocrit = NULL,  Krbc2pu = NULL,  Funbound.plasma = NULL,  default.to.human = FALSE,  species = "Human",  adjusted.Funbound.plasma = TRUE,  class.exclude = TRUE,  suppress.messages = TRUE)

Arguments

Details

The red blood cell (RBC) parition coefficient as predicted by the Schmitt(2008) method is used in the calculation. The value is calculated with theequation: 1 - hematocrit + hematocrit * Krbc2pu * Funbound.plasma, summingthe red blood cell to plasma and plasma:plasma (equal to 1) partitioncoefficients multiplied by their respective fractional volumes. Whenspecies is specified as rabbit, dog, or mouse, the function uses theappropriate physiological data (hematocrit and temperature), but substituteshuman fraction unbound and tissue volumes.

Value

The blood to plasma chemical concentration ratio

Author(s)

John Wambaugh and Robert Pearce

References

Schmitt W (2008).“General approach for the calculation of tissue to plasma partition coefficients.”Toxicology in vitro,22(2), 457–467.doi:10.1016/j.tiv.2007.09.010.

Pearce RG, Setzer RW, Davis JL, Wambaugh JF (2017).“Evaluation and calibration of high-throughput predictions of chemical distribution to tissues.”Journal of pharmacokinetics and pharmacodynamics,44, 549–565.doi:10.1007/s10928-017-9548-7.

Ruark CD, Hack CE, Robinson PJ, Mahle DA, Gearhart JM (2014).“Predicting passive and active tissue: plasma partition coefficients: interindividual and interspecies variability.”Journal of pharmaceutical sciences,103(7), 2189–2198.doi:10.1002/jps.24011.

Examples

calc_rblood2plasma(chem.name="Bisphenol A")calc_rblood2plasma(chem.name="Bisphenol A",species="Rat")

Calculate toxicokinetic summary statistics (deprecated).

Description

#' This function is included for backward compatibility. It callscalc_tkstats which calculates the area under the curve, the mean, and the peak valuesfor the venous blood or plasma concentration of a specified chemical or allchemicals if none is specified for the multiple compartment model with agiven number of days, dose, and number of doses per day.

Usage

calc_stats(  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  parameters = NULL,  route = "oral",  stats = c("AUC", "peak", "mean"),  species = "Human",  days = 28,  daily.dose = 1,  dose = NULL,  doses.per.day = 1,  output.units = "uM",  concentration = "plasma",  tissue = "plasma",  model = "pbtk",  default.to.human = FALSE,  adjusted.Funbound.plasma = TRUE,  regression = TRUE,  restrictive.clearance = TRUE,  suppress.messages = FALSE,  ...)

Arguments

Details

Default value of 0 for doses.per.day solves for a single dose.

When species is specified as rabbit, dog, or mouse, the function uses theappropriate physiological data(volumes and flows) but substitues humanfraction unbound, partition coefficients, and intrinsic hepatic clearance.

Value

Author(s)

Robert Pearce and John Wambaugh

Calculate toxicokinetic summary statistics.

Description

This function calculates the area under the curve, the mean, and the peak valuesfor the venous blood or plasma concentration of a specified chemical or allchemicals if none is specified for the multiple compartment model with agiven number of days, dose, and number of doses per day.

Usage

calc_tkstats(  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  parameters = NULL,  route = "oral",  stats = c("AUC", "peak", "mean"),  species = "Human",  days = 28,  daily.dose = 1,  dose = NULL,  forcings = NULL,  doses.per.day = 1,  output.units = "uM",  concentration = "plasma",  tissue = "plasma",  model = "pbtk",  suppress.messages = FALSE,  ...)

Arguments

Details

Default value of 0 for doses.per.day solves for a single dose.

When species is specified as rabbit, dog, or mouse, the function uses theappropriate physiological data(volumes and flows) but substitues humanfraction unbound, partition coefficients, and intrinsic hepatic clearance.

Value

Author(s)

Robert Pearce and John Wambaugh

Examples

calc_tkstats(chem.name='Bisphenol-A',days=100,stats='mean',model='3compartment')calc_tkstats(chem.name='Bisphenol-A',days=100,stats=c('peak','mean'),species='Rat')triclosan.stats <- calc_tkstats(days=10, chem.name = "triclosan")calc_tkstats(dtxsid="DTXSID0020442",days=1)calc_tkstats(dtxsid="DTXSID0020442",days=10)calc_tkstats(dtxsid="DTXSID0020442",days=100)

Calculate the total plasma clearance.

Description

This function calculates the total clearance rate for a one compartment modelfor plasmawhere clearance is entirely due to metablism by the liver and glomerularfiltration in the kidneys, identical to clearance of three compartmentsteady state model.

Usage

calc_total_clearance(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  model = "3compartmentss",  suppress.messages = FALSE,  species = "Human",  ...)

Arguments

Value

Author(s)

John Wambaugh

Examples

calc_total_clearance(chem.name="Ibuprofen")

Calculate the volume of distribution for a one compartment model.

Description

This function predicts partition coefficients for all tissues usingpredict_partitioning_schmitt, then lumps theminto a single compartment.

Usage

calc_vdist(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  suppress.messages = FALSE,  adjusted.Funbound.plasma = TRUE,  species = "Human",  default.to.human = FALSE,  ...)

Arguments

Details

The effective volume of distribution is calculated by summing each tissuesvolume times it's partition coefficient relative to plasma. Plasma, and theparitioning into RBCs are also added to get the total volume of distributionin L/KG BW. Partition coefficients are calculated using Schmitt's (2008)method. When species is specified as rabbit, dog, or mouse, the functionuses the appropriate physiological data(volumes and flows) but substitueshuman fraction unbound, partition coefficients, and intrinsic hepaticclearance.

Value

Author(s)

John Wambaugh and Robert Pearce

References

Schmitt W (2008).“General approach for the calculation of tissue to plasma partition coefficients.”Toxicology in vitro,22(2), 457–467.doi:10.1016/j.tiv.2007.09.010.

Peyret T, Poulin P, Krishnan K (2010).“A unified algorithm for predicting partition coefficients for PBPK modeling of drugs and environmental chemicals.”Toxicology and applied pharmacology,249(3), 197–207.doi:10.1016/j.taap.2010.09.010.

See Also

predict_partitioning_schmitt

tissue.data

physiology.data

Examples

calc_vdist(chem.cas="80-05-7")calc_vdist(chem.name="Bisphenol A")calc_vdist(chem.name="Bisphenol A",species="Rat")# Create a list of parameters (that you can potentially change):p <- parameterize_schmitt(chem.name="propranolol")# Need to use those parameters to predict partition coefficients:PCs <- predict_partitioning_schmitt(parameters = p)# Lump the tissues into a single volume of distributioncalc_vdist(parameters=c(p,PCs))# Should be the same as chemical by name:calc_vdist(chem.name="propranolol")# Different ways to give the arguments:calc_vdist(chem.cas="80-05-7")params <- parameterize_schmitt(chem.name="triclosan")params <- c(params, predict_partitioning_schmitt(parameters = params))calc_vdist(parameters=params)params <- parameterize_3comp(chem.name="triclosan")calc_vdist(parameters=params)params <- parameterize_pbtk(chem.name="triclosan")calc_vdist(parameters=params)

CAS number format check function

Description

This function checks whether the CAS/CARN chemical identifier follows the anticipatedformat of XXXXXXX-YY-Z (i.e. 2-7 digits, 2 digits, and 1 digit, respectively).

Usage

cas_id_check(cas)

Arguments

Value

Logical output (TRUE or FALSE) indicating whether the character string(s)provided match the anticipated format for a CAS/CASRN chemical identifier.

Check for sufficient model parameters

Description

This function halt model evaluation if not all the needed parameters(as specified in the modelinfo_[MODEL].r file) are available. The function usesget_cheminfo, so if the chemical has been checked against thatfunction already then evaluation should proceed as expected. If you do nothave the parameters you need and are using a non-humanspecies try default.to.human = TRUE (there are many more values for human thanany other species). If working in human, try first usingload_dawson2021,load_sipes2017, orload_pradeep2020.

Usage

check_model(  chem.name = NULL,  chem.cas = NULL,  dtxsid = NULL,  model = NULL,  species = NULL,  class.exclude = TRUE,  physchem.exclude = TRUE,  default.to.human = FALSE)

Arguments

Value

Stops code from running if all parameters needed for modelare not available, otherwise does nothing.

Author(s)

john Wambaugh

See Also

get_cheminfo

Parameter Estimates from Wambaugh et al. (2018)

Description

This table includes 1 and 2 compartment fits of plasma concentration vs timedata aggregated from chem.invivo.PK.data, performed in Wambaugh et al. 2018.Data includes volume of distribution (Vdist, L/kg), elimination rate (kelim,1/h), gut absorption rate (kgutabs, 1/h), fraction absorbed (Fabsgut), andsteady state concentration (Css, mg/L).

Usage

chem.invivo.PK.aggregate.data

Format

data.frame

Author(s)

John Wambaugh

Source

Wambaugh et al. 2018

References

Wambaugh JF, Hughes MF, Ring CL, MacMillan DK, Ford J, Fennell TR, Black SR, Snyder RW, Sipes NS, Wetmore BA, others (2018).“Evaluating in vitro-in vivo extrapolation of toxicokinetics.”Toxicological Sciences,163(1), 152–169.doi:10.1093/toxsci/kfy020.

Summary of published toxicokinetic time course experiments

Description

This data set summarizes the time course data in the chem.invivo.PK.datatable. Maximum concentration (Cmax), time integrated plasma concentrationfor the duration of treatment (AUC.treatment) and extrapolated to zeroconcentration (AUC.infinity) as well as half-life are calculated. Summaryvalues are given for each study and dosage. These data can be used toevaluate toxicokinetic model predictions.

Usage

chem.invivo.PK.summary.data

Format

A data.frame containing 100 rows and 25 columns.

Author(s)

John Wambaugh

Source

Wambaugh et al. 2018 Toxicological Sciences, in press

References

Aanderud L, Bakke OM (1983). Pharmacokinetics of antipyrine,paracetamol, and morphine in rat at 71 ATA. Undersea Biomed Res.10(3):193-201. PMID: 6636344

Aasmoe L, Mathiesen M, Sager G (1999). Elimination of methoxyacetic acid andethoxyacetic acid in rat. Xenobiotica. 29(4):417-24. PMID: 10375010

Ako RA. Pharmacokinetics/pharmacodynamics (PK/PD) of oral diethylstilbestrol(DES) in recurrent prostate cancer patients and of oral dissolving film(ODF)-DES in rats. PhD dissertation, College of Pharmacy, University ofHouston, USA, 2011.

Anadon A, Martinez-Larranaga MR, Fernandez-Cruz ML, Diaz MJ, Fernandez MC,Martinez MA (1996). Toxicokinetics of deltamethrin and its 4'-HO-metabolitein the rat. Toxicol Appl Pharmacol. 141(1):8-16. PMID: 8917670

Binkerd PE, Rowland JM, Nau H, Hendrickx AG (1988). Evaluation of valproicacid (VPA) developmental toxicity and pharmacokinetics in Sprague-Dawleyrats. Fundam Appl Toxicol. 11(3):485-93. PMID: 3146521

Boralli VB, Coelho EB, Cerqueira PM, Lanchote VL (2005). Stereoselectiveanalysis of metoprolol and its metabolites in rat plasma with application tooxidative metabolism. J Chromatogr B Analyt Technol Biomed Life Sci.823(2):195-202. PMID: 16029965

Chan MP, Morisawa S, Nakayama A, Kawamoto Y, Sugimoto M, Yoneda M (2005).Toxicokinetics of 14C-endosulfan in male Sprague-Dawley rats following oraladministration of single or repeated doses. Environ Toxicol. 20(5):533-41.PMID: 16161119

Cruz L, Castaneda-Hernandez G, Flores-Murrieta FJ, Garcia-Lopez P,Guizar-Sahagun G (2002). Alteration of phenacetin pharmacokinetics afterexperimental spinal cord injury. Proc West Pharmacol Soc. 45:4-5. PMID:12434508

Della Paschoa OE, Mandema JW, Voskuyl RA, Danhof M (1998).Pharmacokinetic-pharmacodynamic modeling of the anticonvulsant andelectroencephalogram effects of phenytoin in rats. J Pharmacol Exp Ther.284(2):460-6. PMID: 9454785

Du B, Li X, Yu Q, A Y, Chen C (2010). Pharmacokinetic comparison of orallydisintegrating, beta-cyclodextrin inclusion complex and conventional tabletsof nicardipine in rats. Life Sci J. 7(2):80-4.

Farris FF, Dedrick RL, Allen PV, Smith JC (1993). Physiological model forthe pharmacokinetics of methyl mercury in the growing rat. Toxicol ApplPharmacol. 119(1):74-90. PMID: 8470126

Hays SM, Elswick BA, Blumenthal GM, Welsch F, Conolly RB, Gargas ML (2000).Development of a physiologically based pharmacokinetic model of2-methoxyethanol and 2-methoxyacetic acid disposition in pregnant rats.Toxicol Appl Pharmacol. 163(1):67-74. PMID: 10662606

Igari Y, Sugiyama Y, Awazu S, Hanano M (1982). Comparative physiologicallybased pharmacokinetics of hexobarbital, phenobarbital and thiopental in therat. J Pharmacokinet Biopharm. 10(1):53-75. PMID: 7069578

Ito K, Houston JB (2004). Comparison of the use of liver models forpredicting drug clearance using in vitro kinetic data from hepaticmicrosomes and isolated hepatocytes. Pharm Res. 21(5):785-92. PMID: 15180335

Jia L, Wong H, Wang Y, Garza M, Weitman SD (2003). Carbendazim: disposition,cellular permeability, metabolite identification, and pharmacokineticcomparison with its nanoparticle. J Pharm Sci. 92(1):161-72. PMID: 12486692

Kawai R, Mathew D, Tanaka C, Rowland M (1998). Physiologically basedpharmacokinetics of cyclosporine A: extension to tissue distributionkinetics in rats and scale-up to human. J Pharmacol Exp Ther. 287(2):457-68.PMID: 9808668

Kim YC, Kang HE, Lee MG (2008). Pharmacokinetics of phenytoin and itsmetabolite, 4'-HPPH, after intravenous and oral administration of phenytointo diabetic rats induced by alloxan or streptozotocin. Biopharm Drug Dispos.29(1):51-61. PMID: 18022993

Kobayashi S, Takai K, Iga T, Hanano M (1991). Pharmacokinetic analysis ofthe disposition of valproate in pregnant rats. Drug Metab Dispos.19(5):972-6. PMID: 1686245

Kotegawa T, Laurijssens BE, Von Moltke LL, Cotreau MM, Perloff MD,Venkatakrishnan K, Warrington JS, Granda BW, Harmatz JS, Greenblatt DJ(2002). In vitro, pharmacokinetic, and pharmacodynamic interactions ofketoconazole and midazolam in the rat. J Pharmacol Exp Ther. 302(3):1228-37.PMID: 12183684

Krug AK, Kolde R, Gaspar JA, Rempel E, Balmer NV, Meganathan K, Vojnits K,Baquie M, Waldmann T, Ensenat-Waser R, Jagtap S, Evans RM, Julien S,Peterson H, Zagoura D, Kadereit S, Gerhard D, Sotiriadou I, Heke M,Natarajan K, Henry M, Winkler J, Marchan R, Stoppini L, Bosgra S, WesterhoutJ, Verwei M, Vilo J, Kortenkamp A, Hescheler J, Hothorn L, Bremer S, vanThriel C, Krause KH, Hengstler JG, Rahnenfuhrer J, Leist M, Sachinidis A(2013). Human embryonic stem cell-derived test systems for developmentalneurotoxicity: a transcriptomics approach. Arch Toxicol. 87(1):123-43. PMID:23179753

Leon-Reyes MR, Castaneda-Hernandez G, Ortiz MI (2009). Pharmacokinetic ofdiclofenac in the presence and absence of glibenclamide in the rat. J PharmPharm Sci. 12(3):280-7. PMID: 20067705

Nagata M, Hidaka M, Sekiya H, Kawano Y, Yamasaki K, Okumura M, Arimori K(2007). Effects of pomegranate juice on human cytochrome P450 2C9 andtolbutamide pharmacokinetics in rats. Drug Metab Dispos. 35(2):302-5. PMID:17132763

Okiyama M, Ueno K, Ohmori S, Igarashi T, Kitagawa H (1988). Druginteractions between imipramine and benzodiazepines in rats. J Pharm Sci.77(1):56-63. PMID: 2894451

Pelissier-Alicot AL, Schreiber-Deturmeny E, Simon N, Gantenbein M,Bruguerolle B (2002). Time-of-day dependent pharmacodynamic andpharmacokinetic profiles of caffeine in rats. Naunyn Schmiedebergs ArchPharmacol. 365(4):318-25. PMID: 11919657

Piersma AH, Bosgra S, van Duursen MB, Hermsen SA, Jonker LR, Kroese ED, vander Linden SC, Man H, Roelofs MJ, Schulpen SH, Schwarz M, Uibel F, vanVugt-Lussenburg BM, Westerhout J, Wolterbeek AP, van der Burg B (2013).Evaluation of an alternative in vitro test battery for detectingreproductive toxicants. Reprod Toxicol. 38:53-64. PMID: 23511061

Pollack GM, Li RC, Ermer JC, Shen DD (1985). Effects of route ofadministration and repetitive dosing on the disposition kinetics ofdi(2-ethylhexyl) phthalate and its mono-de-esterified metabolite in rats.Toxicol Appl Pharmacol. Jun 30;79(2):246-56. PMID: 4002226

Saadeddin A, Torres-Molina F, Carcel-Trullols J, Araico A, Peris JE (2004).Pharmacokinetics of the time-dependent elimination of all-trans-retinoicacid in rats. AAPS J. 6(1):1-9. PMID: 18465253

Satterwhite JH, Boudinot FD (1991). Effects of age and dose on thepharmacokinetics of ibuprofen in the rat. Drug Metab Dispos. 19(1):61-7.PMID: 1673423

Szymura-Oleksiak J, Panas M, Chrusciel W (1983). Pharmacokinetics ofimipramine after single and multiple intravenous administration in rats. PolJ Pharmacol Pharm. 35(2):151-7. PMID: 6622297

Tanaka C, Kawai R, Rowland M (2000). Dose-dependent pharmacokinetics ofcyclosporin A in rats: events in tissues. Drug Metab Dispos. 28(5):582-9.PMID: 10772639

Timchalk C, Nolan RJ, Mendrala AL, Dittenber DA, Brzak KA, Mattsson JL(2002). A Physiologically based pharmacokinetic and pharmacodynamic(PBPK/PD) model for the organophosphate insecticide chlorpyrifos in rats andhumans. Toxicol Sci. Mar;66(1):34-53. PMID: 11861971

Tokuma Y, Sekiguchi M, Niwa T, Noguchi H (1988). Pharmacokinetics ofnilvadipine, a new dihydropyridine calcium antagonist, in mice, rats,rabbits and dogs. Xenobiotica 18(1):21-8. PMID: 3354229

Treiber A, Schneiter R, Delahaye S, Clozel M (2004). Inhibition of organicanion transporting polypeptide-mediated hepatic uptake is the majordeterminant in the pharmacokinetic interaction between bosentan andcyclosporin A in the rat. J Pharmacol Exp Ther. 308(3):1121-9. PMID:14617681

Tsui BC, Feng JD, Buckley SJ, Yeung PK (1994). Pharmacokinetics andmetabolism of diltiazem in rats following a single intra-arterial or singleoral dose. Eur J Drug Metab Pharmacokinet. 19(4):369-73. PMID: 7737239

Wambaugh, John F., et al. "Toxicokinetic triage for environmentalchemicals." Toxicological Sciences (2015): 228-237.

Wang Y, Roy A, Sun L, Lau CE (1999). A double-peak phenomenon in thepharmacokinetics of alprazolam after oral administration. Drug Metab Dispos.27(8):855-9. PMID: 10421610

Wang X, Lee WY, Or PM, Yeung JH (2010). Pharmacokinetic interaction studiesof tanshinones with tolbutamide, a model CYP2C11 probe substrate, usingliver microsomes, primary hepatocytes and in vivo in the rat. Phytomedicine.17(3-4):203-11. PMID: 19679455

Yang SH, Lee MG (2008). Dose-independent pharmacokinetics of ondansetron inrats: contribution of hepatic and intestinal first-pass effects to lowbioavailability. Biopharm Drug Dispos. 29(7):414-26. PMID: 18697186

Yeung PK, Alcos A, Tang J (2009). Pharmacokinetics and Hemodynamic Effectsof Diltiazem in Rats Following Single vs Multiple Doses In Vivo. Open DrugMetab J. 3:56-62.

Physico-chemical properties and in vitro measurements for toxicokinetics

Description

This data set contains the necessary information to make basic,high-throughput toxicokinetic (HTTK) predictions for compounds, includingFunbound.plasma, molecular weight (g/mol), logP, logMA (membrane affinity),intrinsic clearance(uL/min/10^6 cells), and pKa. These data have beencompiled from multiple sources, and can be used to parameterize a variety oftoxicokinetic models. See variable EPA.ref for information on the reference EPA.

Usage

chem.physical_and_invitro.data

Format

A data.frame containing 9411 rows and 54 columns.

Details

In some cases the rapid equilbrium dailysis method (Waters et al., 2008)fails to yield detectable concentrations for the free fraction of chemical. In those cases we assume the compound is highly bound (that is, Fup approacheszero). For some calculations (for example, steady-state plasma concentration)there is precendent (Rotroff et al., 2010) for using half the average limit of detection, that is 0.005. We do not recomend using other models where quantities like partition coefficients must be predicted using Fup. We alsodo not recomend including the value 0.005 in training sets for Fup predictivemodels.

Note that in some cases theFunbound.plasma and theintrinsic clearance areprovided as a series of numbers separated by commas. These values are the result of Bayesian analysis and characterize a distribution: the first valueis the median of the distribution, while the second and third values are the lower and upper 95th percentile (that is qunatile 2.5 and 97.5) respectively.For intrinsic clearance a fourth value indicating a p-value for a decrease isprovided. Typically 4000 samples were used for the Bayesian analusis, suchthat a p-value of "0" is equivale to "<0.00025". See Wambaugh et al. (2019)for more details.

Any one chemical compoundmay have multiple ionization equilibria (see Strope et al., 2018) may both for donating or accepting a proton (andtherefore changing charge state). If there are multiple equlibria of the sametype (donor/accept])the are concatonated by commas.

All species-specific information is initially from experimental measurements.The functionsload_sipes2017,load_pradeep2020, andload_dawson2021 may be used to add in silico, structure-basedpredictions for many thousands of additional compounds to this table.

Author(s)

John Wambaugh

Source

Wambaugh, John F., et al. "Toxicokinetic triage for environmentalchemicals." Toxicological Sciences (2015): 228-237.

References

CompTox Chemicals Dashboard (https://comptox.epa.gov/dashboard)

EPI Suite, https://www.epa.gov/opptintr/exposure/pubs/episuite.htm

Brown, Hayley S., Michael Griffin, and J. Brian Houston. "Evaluation of cryopreserved human hepatocytes as an alternative in vitro system to microsomes for the prediction of metabolic clearance." Drug metabolism and disposition 35.2 (2007): 293-301.

Gulden, Michael, et al. "Impact of protein binding on the availability and cytotoxic potency of organochlorine pesticides and chlorophenols in vitro." Toxicology 175.1-3 (2002): 201-213.

Hilal, S., Karickhoff, S. and Carreira, L. (1995). A rigorous test forSPARC's chemical reactivity models: Estimation of more than 4300 ionizationpKas. Quantitative Structure-Activity Relationships 14(4), 348-355.

Honda GS, Pearce RG, Pham LL, Setzer RW, Wetmore BA, Sipes NS, Gilbert J, Franz B, Thomas RS, Wambaugh JF (2019).“Using the concordance of in vitro and in vivo data to evaluate extrapolation assumptions.”PloS one,14(5), e0217564.doi:10.1371/journal.pone.0217564.

Ito K, Houston JB (2004).“Comparison of the use of liver models for predicting drug clearance using in vitro kinetic data from hepatic microsomes and isolated hepatocytes.”Pharmaceutical research,21, 785–792.doi:10.1023/B:PHAM.0000026429.12114.7d.

Jones, O. A., Voulvoulis, N. and Lester, J. N. (2002). Aquatic environmentalassessment of the top 25 English prescription pharmaceuticals. Waterresearch 36(20), 5013-22.

Jones, Barry C., et al. "An investigation into the prediction of in vivo clearance for a range of flavin-containing monooxygenase substrates." Drug metabolism and disposition 45.10 (2017): 1060-1067.

Lau, Y. Y., Sapidou, E., Cui, X., White, R. E. and Cheng, K. C. (2002).Development of a novel in vitro model to predict hepatic clearance usingfresh, cryopreserved, and sandwich-cultured hepatocytes. Drug Metabolism andDisposition 30(12), 1446-54.

Linakis MW, Sayre RR, Pearce RG, Sfeir MA, Sipes NS, Pangburn HA, Gearhart JM, Wambaugh JF (2020).“Development and evaluation of a high-throughput inhalation model for organic chemicals.”Journal of exposure science & environmental epidemiology,30(5), 866–877.doi:10.1038/s41370-020-0238-y.

Lombardo, F., Berellini, G., & Obach, R. S. (2018). Trend analysis of a database of intravenous pharmacokinetic parameters in humans for 1352 drug compounds. Drug Metabolism and Disposition, 46(11), 1466-1477.

McGinnity, D. F., Soars, M. G., Urbanowicz, R. A. and Riley, R. J. (2004).Evaluation of fresh and cryopreserved hepatocytes as in vitro drugmetabolism tools for the prediction of metabolic clearance. Drug Metabolismand Disposition 32(11), 1247-53, 10.1124/dmd.104.000026.

Naritomi, Y., Terashita, S., Kagayama, A. and Sugiyama, Y. (2003). Utilityof Hepatocytes in Predicting Drug Metabolism: Comparison of HepaticIntrinsic Clearance in Rats and Humans in Vivo and in Vitro. Drug Metabolismand Disposition 31(5), 580-588, 10.1124/dmd.31.5.580.

Obach, R. S. (1999). Prediction of human clearance of twenty-nine drugs fromhepatic microsomal intrinsic clearance data: An examination of in vitrohalf-life approach and nonspecific binding to microsomes. Drug Metabolismand Disposition 27(11), 1350-9.

Paini, Alicia; Cole, Thomas; Meinero, Maria; Carpi, Donatella; Deceuninck, Pierre; Macko, Peter; Palosaari, Taina; Sund, Jukka; Worth, Andrew; Whelan, Maurice (2020): EURL ECVAM in vitro hepatocyte clearance and blood plasma protein binding dataset for 77 chemicals. European Commission, Joint Research Centre (JRC) [Dataset] PID: https://data.europa.eu/89h/a2ff867f-db80-4acf-8e5c-e45502713bee

Paixao, P., Gouveia, L. F., & Morais, J. A. (2012). Prediction of the humanoral bioavailability by using in vitro and in silico drug related parametersin a physiologically based absorption model. International journal ofpharmaceutics, 429(1), 84-98.

Pirovano, Alessandra, et al. "QSARs for estimating intrinsic hepaticclearance of organic chemicals in humans." Environmental toxicology andpharmacology 42 (2016): 190-197.

Riley, Robert J., Dermot F. McGinnity, and Rupert P. Austin. "A unified model for predicting human hepatic, metabolic clearance from in vitro intrinsic clearance data in hepatocytes and microsomes." Drug Metabolism and Disposition 33.9 (2005): 1304-1311.

Schmitt W (2008).“General approach for the calculation of tissue to plasma partition coefficients.”Toxicology in vitro,22(2), 457–467.doi:10.1016/j.tiv.2007.09.010.

Shibata, Y., Takahashi, H., Chiba, M. and Ishii, Y. (2002). Prediction ofHepatic Clearance and Availability by Cryopreserved Human Hepatocytes: AnApplication of Serum Incubation Method. Drug Metabolism and Disposition30(8), 892-896, 10.1124/dmd.30.8.892.

Sohlenius-Sternbeck, Anna-Karin, et al. "Practical use of the regression offset approach for the prediction of in vivo intrinsic clearance from hepatocytes." Xenobiotica 42.9 (2012): 841-853.

Tonnelier, A., Coecke, S. and Zaldivar, J.-M. (2012). Screening of chemicalsfor human bioaccumulative potential with a physiologically basedtoxicokinetic model. Archives of Toxicology 86(3), 393-403,10.1007/s00204-011-0768-0.

Uchimura, Takahide, et al. "Prediction of human blood-to-plasma drugconcentration ratio." Biopharmaceutics & drug disposition 31.5-6 (2010):286-297.

Wambaugh JF, Wetmore BA, Ring CL, Nicolas CI, Pearce RG, Honda GS, Dinallo R, Angus D, Gilbert J, Sierra T, others (2019).“Assessing toxicokinetic uncertainty and variability in risk prioritization.”Toxicological Sciences,172(2), 235–251.doi:10.1093/toxsci/kfz205.

Wetmore BA, Wambaugh JF, Ferguson SS, Sochaski MA, Rotroff DM, Freeman K, Clewell III HJ, Dix DJ, Andersen ME, Houck KA, others (2012).“Integration of dosimetry, exposure, and high-throughput screening data in chemical toxicity assessment.”Toxicological Sciences,125(1), 157–174.doi:10.1093/toxsci/kfr254.

Wetmore BA, Wambaugh JF, Ferguson SS, Li L, Clewell III HJ, Judson RS, Freeman K, Bao W, Sochaski MA, Chu T, others (2013).“Relative impact of incorporating pharmacokinetics on predicting in vivo hazard and mode of action from high-throughput in vitro toxicity assays.”toxicological sciences,132(2), 327–346.doi:10.1093/toxsci/kft012.

Wetmore BA, Wambaugh JF, Allen B, Ferguson SS, Sochaski MA, Setzer RW, Houck KA, Strope CL, Cantwell K, Judson RS, others (2015).“Incorporating high-throughput exposure predictions with dosimetry-adjusted in vitro bioactivity to inform chemical toxicity testing.”Toxicological Sciences,148(1), 121–136.doi:10.1093/toxsci/kfv171.

F. L. Wood, J. B. Houston and D. Hallifax'Drug Metabolism and Disposition November 1, 2017, 45 (11) 1178-1188; DOI: https://doi.org/10.1124/dmd.117.077040

See Also

get_physchem_param

get_invitroPK_param

add_chemtable

CKD-EPI equation for GFR.

Description

Predict GFR from serum creatinine, gender, and age.

Usage

ckd_epi_eq(scr, gender, reth, age_years, ckd_epi_race_coeff = FALSE)

Arguments

Details

From Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF, Feldman HI, et al. A newequation to estimate glomerular filtration rate. Ann Intern Med 2009;150(9):604-612. doi:10.7326/0003-4819-150-9-200905050-00006

Value

Vector of GFR values in mL/min/1.73m^2.

Author(s)

Caroline Ring

References

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

convert_solve_x

Description

This function is designed to convert compartment values estimated from oneof the HTTK models (e.g. "1compartment) using the solve_model function.It takes the HTTK model output matrix, model name, desired output units,and compound information to perform the conversion default model units touser specified units.

Usage

convert_solve_x(  model.output.mat,  model = NULL,  output.units = NULL,  MW = NULL,  vol = NULL,  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  monitor.vars = NULL,  suppress.messages = FALSE,  verbose = FALSE,  ...)

Arguments

Details

The function can be used to convert all compartments to a single unit,only units for a single model compartment, or units for a set of modelcompartments.

More details on the unit conversion can be found in the documentation forconvert_units.

Value

'new.ouput.matrix' A matrix with a column for time (in days), eachcompartment, and the area under the curve (AUC) and a rowfor each time point. The compartment and AUC columns areconverted from model specified units to user specified units.

'output.units.vector' A vector of character strings providing themodel compartments and their corresponding units afterconvert_solve_x.

Author(s)

Sarah E. Davidson

See Also

convert_units

Examples

output.mat <- solve_1comp(dtxsid = "DTXSID0020573",days=1)new.output.mat <- convert_solve_x(output.units = "mg",                                  model.output.mat = output.mat,                                  model = "1compartment",                                  dtxsid = "DTXSID0020573")

convert_units

Description

This function is designed to accept input units, output units, and the molecular weight (MW) of a substance of interest to then use a table lookupto return a scaling factor that can be readily applied for the intendedconversion. It can also take chemical identifiers in the place of a specified molecular weight value to retrieve that value for its own use.

Usage

convert_units(  input.units = NULL,  output.units = NULL,  MW = NULL,  vol = NULL,  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  temp = 25,  liquid.density = 1,  state = "liquid")

Arguments

Details

If input or output units not contained in the table are queried,it gives a corresponding error message. It gives a warning message about thehandling of 'ppmv,' as the function is only set up to convert between ppmv and mass-based units (like mg/m³ or umol/L) in the context of ideal gases.

convert_units is not directly configured to accept and convert units basedon BW, like mg/kg. For this purpose, seescale_dosing.

The function supports a limited set of most relevant units acrosstoxicological models, currently including umol, uM, mg, mg/L, mg/m³ or umol/L), andin the context of gases assumed to be ideal, ppmv.

Andersen and Clewell's Rules of PBPK Modeling:

1. Check Your Units

2. Check Your Units

3. Check Mass Balance

Author(s)

Mark Sfeir, John Wambaugh, and Sarah E. Davidson

Examples

# MW BPA is 228.29 g/mol# 1 mg/L -> 1/228.29*1000 = 4.38 uMconvert_units("mg/L","uM",chem.cas="80-05-7")# MW Diclofenac is 296.148 g/mol# 1 uM -> 296.148/1000 =  0.296convert_units("uM","mg/L",chem.name="diclofenac")# ppmv only works for gasses:try(convert_units("uM","ppmv",chem.name="styrene"))convert_units("uM","ppmv",chem.name="styrene",state="gas")# Compare with https://www3.epa.gov/ceampubl/learn2model/part-two/onsite/ia_unit_conversion.html# 1 ug/L Toluene -> 0.263 ppmvconvert_units("ug/L","ppmv",chem.name="toluene",state="gas")# 1 pppmv Toluene, 0.0038 mg/Lconvert_units("ppmv","mg/L",chem.name="toluene",state="gas")MW_pyrene <- get_physchem_param(param='MW', chem.name='pyrene')conversion_factor <- convert_units(input.units='mg/L', output.units ='uM',  MW=MW_pyrene)calc_mc_oral_equiv(15, parameters=p)

Create a table of parameter values for Monte Carlo

Description

This is the HTTK master function for creating a data table foruse with Monte Carlo methods to simulate parameter uncertainty andvariabilit. Each column of the output table corresponds to an HTTK modelparameter and each row corresponds to a different random draw (for example,different individuals when considering biological variability). Thisfunction call three different key functions to simulate parameter parameteruncertainty and/or variability in one of three ways. First parameters canbe varied in an uncorrelated manner using truncated normal distributions bythe functionmonte_carlo. Then, physiological parameters canbe varied in a correlated manner according to theRing et al. (2017)(doi:10.1016/j.envint.2017.06.004)httk-pop approach by thefunctionhttkpop_mc. Next, both uncertainty and variabilityof in vitro HTTK parameters can be simulated by the functioninvitro_mc as described byWambaugh et al. (2019)(doi:10.1093/toxsci/kfz205). Finally, tissue-specific partitioncoefficients are predicted for each draw using theSchmitt (2008)(doi:10.1016/j.tiv.2007.09.010) method as calibrated toin vivodata by Pearce et al. (2017)(doi:10.1007/s10928-017-9548-7) and implemented inpredict_partitioning_schmitt.

Usage

create_mc_samples(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  parameters = NULL,  samples = 1000,  species = "Human",  suppress.messages = FALSE,  model = "3compartmentss",  httkpop = TRUE,  invitrouv = TRUE,  calcrb2p = TRUE,  censored.params = list(),  vary.params = list(),  return.samples = FALSE,  tissue = NULL,  httkpop.dt = NULL,  invitro.mc.arg.list = NULL,  adjusted.Funbound.plasma = NA,  adjusted.Clint = NA,  httkpop.generate.arg.list = list(method = "direct resampling"),  convert.httkpop.arg.list = NULL,  propagate.invitrouv.arg.list = NULL,  parameterize.args.list = NULL,  Caco2.options = NULL)

Arguments

Details

The Monte Carlo methods used here were recently updated anddescribed by Breen et al. (2022).

We aim to make any function that uses chemical identifiers (name, CAS,DTXSID) also work if passed a complete vector of parameters (that is, a rowfrom the table generated by this function). This allows the use of MonteCarlo to vary the parameters and therefore vary the function output.Depending on the type of parameters (for example, physiological vs. invitro measurements) we vary the parameters in different ways with differentfunctions.

NOTE: This function calculates oral bioavailability parameters based onCaco2 data. If you are comparing the output of this function to the outputof a model-parameterization function 'parameterize_MODEL()', you will needto to ensure that 'parameterize_MODEL()' also uses Caco2 data. The built-inmodel parameterization functions default to using *in vivo* oralbioavailability data when available. To force them to use Caco2 datainstead, please pass the following argument to 'parameterize_MODEL()':'Caco2.options = list(overwrite.invivo = TRUE'.

Value

A data table where each column corresponds to parameters needed forthe specified model and each row represents a different Monte Carlo sampleof parameter values.

Author(s)

Caroline Ring, Robert Pearce, and John Wambaugh

References

Breen M, Wambaugh JF, Bernstein A, Sfeir M, Ring CL (2022).“Simulating toxicokinetic variability to identify susceptible and highly exposed populations.”Journal of Exposure Science & Environmental Epidemiology,32(6), 855–863.doi:10.1038/s41370-022-00491-0.

Pearce RG, Setzer RW, Davis JL, Wambaugh JF (2017).“Evaluation and calibration of high-throughput predictions of chemical distribution to tissues.”Journal of pharmacokinetics and pharmacodynamics,44, 549–565.doi:10.1007/s10928-017-9548-7.

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

Schmitt W (2008).“General approach for the calculation of tissue to plasma partition coefficients.”Toxicology in vitro,22(2), 457–467.doi:10.1016/j.tiv.2007.09.010.

Wambaugh JF, Wetmore BA, Ring CL, Nicolas CI, Pearce RG, Honda GS, Dinallo R, Angus D, Gilbert J, Sierra T, others (2019).“Assessing toxicokinetic uncertainty and variability in risk prioritization.”Toxicological Sciences,172(2), 235–251.doi:10.1093/toxsci/kfz205.

Examples

# We can use the Monte Carlo functions by passing a table# where each row represents a different Monte Carlo draw of parameters:p <- create_mc_samples(chem.cas="80-05-7")# Use data.table for steady-state plasma concentration (Css) Monte Carlo:calc_mc_css(parameters=p)# Using the same table gives the same answer:calc_mc_css(parameters=p)# Use Css for 1 mg/kg/day for simple reverse toxicokinetics# in Vitro-In Vivo Extrapolation to convert 15 uM to mg/kg/day:15/calc_mc_css(parameters=p, output.units="uM")# Can do the same with calc_mc_oral_equiv:calc_mc_oral_equiv(15, parameters=p)#Generate a population using the virtual-individuals method,#including 80 females and 20 males,#including only ages 20-65,#including only Mexican American and#Non-Hispanic Black individuals,#including only non-obese individualsset.seed(42)mypop <- httkpop_generate(method = 'virtual individuals',                          gendernum=list(Female=80,                          Male=20),                          agelim_years=c(20,65),                          reths=c('Mexican American',                          'Non-Hispanic Black'),                          weight_category=c('Underweight',                          'Normal',                          'Overweight'))# Including a httkpop.dt argument will overwrite the number of sample and# the httkpop on/off logical switch:samps1 <- create_mc_samples(chem.name="bisphenola",                           httkpop=FALSE,                           httkpop.dt=mypop)samps2 <- create_mc_samples(chem.name="bisphenola",                           httkpop.dt=mypop)# But we can turn httkpop off altogether if desired:samps3 <- create_mc_samples(chem.name="bisphenola",                           httkpop=FALSE)

Dawson et al. 2021 data

Description

This table includes QSAR (Random Forest) model predicted values for unboundfraction plasma protein (fup) and intrinsic hepatic clearance (clint) for asubset of chemicals in the Tox21 library(seehttps://www.epa.gov/chemical-research/toxicology-testing-21st-century-tox21).

Usage

dawson2021

Format

data.frame

Details

Predictions were made with a set of Random Forest QSAR models,as reported in Dawson et al. (2021).

Author(s)

Daniel E. Dawson

References

Dawson DE, Ingle BL, Phillips KA, Nichols JW, Wambaugh JF, Tornero-Velez R (2021).“Designing QSARs for Parameters of High-Throughput Toxicokinetic Models Using Open-Source Descriptors.”Environmental Science & Technology,55(9), 6505-6517.doi:10.1021/acs.est.0c06117.PMID: 33856768, https://doi.org/10.1021/acs.est.0c06117.

See Also

load_dawson2021

Machine Learning PFAS Half-Life Predictions from Dawson et al. 2023

Description

Dawson et al. (2023) Supplemental Information S3 includes half-life predictions for 6603 PFAS, of which 3890 are estimated to be within the applicabilitydomain (AD) for humans. This machine learning (ML) model predicts PFAS half-life as one of four categories. The ML model was trained to a dataset of 91 in vivo measured TK half-lives across 11 PFAS, 4 species, and two sexes. Predictions were a function of compound-specific physico-chemical descriptors, species-specific physiological descriptors, and an indicator variable for sex. The kinetics of PFAS are thought to be complicated by active transport, both through either proximal tubular resorption (into the blood) (Andersen et al. 2006) or secretion (into the urine) (Kudo et al. 2002). The ML model uses several species- and structure-derivedsurrogates for estimating the likelihood of active PFAS transport. Geometryof the proximal tubule was a surrogate for transporter expression: sincesecretion/resorption transporters line the surface of the proximal tubule,the amount of surface area provides an upper limit on the amount oftransporter expression. PFAS similarity to three distinct endogenous ligandswas considered as a surrogate for transporter affinity.

Usage

dawson2023

Format

data.frame

Details

The Dawson et al. (2023) half-life categories are:

The data.frame contains the following columns:

References

Dawson DE, Lau C, Pradeep P, Sayre RR, Judson RS, Tornero-Velez R, Wambaugh JF (2023).“A machine learning model to estimate toxicokinetic half-lives of per-and polyfluoro-alkyl substances (PFAS) in multiple species.”Toxics,11(2), 98.

Andersen ME, Clewell III HJ, Tan Y, Butenhoff JL, Olsen GW (2006).“Pharmacokinetic modeling of saturable, renal resorption of perfluoroalkylacids in monkeys?probing the determinants of long plasma half-lives.”Toxicology,227(1-2), 156–164.

Kudo N, Katakura M, Sato Y, Kawashima Y (2002).“Sex hormone-regulated renal transport of perfluorooctanoic acid.”Chemico-biological interactions,139(3), 301–316.

DTXSID number format check function

Description

This function checks whether the DTXSID chemical identifier follows the anticipatedformat of "DTXSID<uniqueID>".

Usage

dtxsid_id_check(dtxsid)

Arguments

Value

Logical output (TRUE or FALSE) indicating whether the character string(s)provided match the anticipated format for a DTXSID chemical identifier.

Predict GFR.

Description

Predict GFR using CKD-EPI equation (for adults) or BSA-based equation (for children).

Usage

estimate_gfr(gfrtmp.dt, gfr_resid_var = TRUE, ckd_epi_race_coeff = FALSE)

Arguments

Details

Add residual variability based on reported residuals for each equation.

Value

The same data.table with agfr_est column added, containing estimated GFR values.

Author(s)

Caroline Ring

References

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

Predict GFR in children.

Description

BSA-based equation from Johnson et al. 2006, Clin Pharmacokinet 45(9) 931-56. Used in Wetmore et al. 2014.

Usage

estimate_gfr_ped(BSA)

Arguments

Value

Vector of GFRs in mL/min/1.73m^2.

Author(s)

Caroline Ring

References

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

Generate hematocrit values for a virtual population

Description

Predict hematocrit from age using smoothing splines and kernel densityestimates of residual variability fitted to NHANES data, for a givencombination of gender and NHANES race/ethnicity category.

Usage

estimate_hematocrit(gender, reth, age_years, age_months, nhanes_mec_svy)

Arguments

Details

This function should usually not be called directly by the user. It is used byhttkpop_generate() in "virtual-individuals" mode, after drawing gender,NHANES race/ethnicity category, and age from their NHANESproportions/distributions.

Value

A vector of numeric generated hematocrit values (blood percentage redblood cells by volume).

Author(s)

Caroline Ring

References

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

SEEM Example DataWe can grab SEEM daily intake rate predictions already in RData format fromhttps://github.com/HumanExposure/SEEM3RPackage/tree/main/SEEM3/dataDownload the file Ring2018Preds.RData

Description

We do not have the space to distribute all the SEEM predictions withinthis R package, but we can give you our "Intro to IVIVE" example chemicals

Usage

example.seem

Format

data.frame

References

Ring CL, Arnot JA, Bennett DH, Egeghy PP, Fantke P, Huang L, Isaacs KK, Jolliet O, Phillips KA, Price PS, others (2018).“Consensus modeling of median chemical intake for the US population based on predictions of exposure pathways.”Environmental science & technology,53(2), 719–732.doi:10.1021/acs.est.8b04056.

ToxCast Example DataThe main page for the ToxCast data is here:https://www.epa.gov/comptox-tools/exploring-toxcast-dataMost useful to us is a single file containing all the hits across all chemcialsand assays:https://clowder.edap-cluster.com/datasets/6364026ee4b04f6bb1409eda?space=62bb560ee4b07abf29f88fef

Description

As of November, 2022 the most recent version was 3.5 and was available as an.Rdata file (invitrodb_3_5_mc5.Rdata)

Usage

example.toxcast

Format

data.frame

Details

Unfortunately for this vignette there are too many ToxCast data to fit into a5mb R package. So we will subset to just the shemicals for the "Intro to IVIVE" vignette and distributeonly those data. In addition, out of 78 columns in the data, we will keep only eight.

Export model to jarnac.

Description

This function exports the multiple compartment PBTK model to a jarnac file.

Usage

export_pbtk_jarnac(  initial.amounts = list(Agutlumen = 0),  folder = tempdir(),  filename = "default.jan",  digits = 4,  ...)

Arguments

Details

Compartments to enter into the initial.amounts list includes Agutlumen,Aart, Aven, Alung, Agut, Aliver, Akidney, and Arest.

Parameters are generated by a call toparameterize_pbtk.When species is specified as rabbit, dog, or mouse, the function uses theappropriate physiological data(volumes and flows) but substitues humanfraction unbound, partition coefficients, and intrinsic hepatic clearance.

Value

Text containing a Jarnac language version of the PBTK model.

Author(s)

Robert Pearce

Examples

export_pbtk_jarnac(chem.name='Nicotine',initial.amounts=list(Agutlumen=1),filename='PBTKmodel.jan')

Export model to sbml.

Description

This function exports the multiple compartment PBTK model to an sbml file.

Usage

export_pbtk_sbml(  initial.amounts = list(Agutlumen = 0),  filename = "default.xml",  folder = tempdir(),  digits = 4,  ...)

Arguments

Details

Compartments to enter into the initial.amounts list includes Agutlumen,Aart, Aven, Alung, Agut, Aliver, Akidney, and Arest.

Parameters are generated by a call toparameterize_pbtk.When species is specified as rabbit, dog, or mouse, the function uses theappropriate physiological data(volumes and flows) but substitues humanfraction unbound, partition coefficients, and intrinsic hepatic clearance.

Value

Text describing the PBTK model in SBML.

Author(s)

Robert Pearce

Examples

export_pbtk_sbml(chem.name='Nicotine',initial.amounts=list(Agutlumen=1),filename='PBTKmodel.xml')

Generate demographic parameters for a virtual population

Description

Generate gender, NHANES race/ethnicity category, ages, heights, and weightsfor a virtual population, based on NHANES data.

Usage

gen_age_height_weight(  nsamp = NULL,  gendernum = NULL,  reths,  weight_category,  agelim_years,  agelim_months,  nhanes_mec_svy)

Arguments

Details

This function should usually not be called directly by the user. It is used byhttkpop_generate() in "virtual-individuals" mode.

Value

A data.table containing variables

gender

Gender of each virtual individual

reth

Race/ethnicity of each virtual individual

age_months

Age in months of each virtual individual

age_years

Age in years of each virtual individual

weight

Body weight in kg of each virtual individual

height

Height in cm of each virtual individual

Author(s)

Caroline Ring

References

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

importFrom survey svymean

Generate heights and weights for a virtual population.

Description

Predict height and weight from age using smoothing splines, and then addresidual variability from a 2-D KDE, both fitted to NHANES data, for a givencombination of gender and NHANES race/ethnicity category.

Usage

gen_height_weight(gender, reth, age_months, nhanes_mec_svy)

Arguments

Details

This function should usually not be called directly by the user. It is used byhttkpop_generate() in "virtual-individuals" mode, after drawing gender,NHANES race/ethnicity category, and age from their NHANESproportions/distributions.

Value

A list containing two named elements,weight andheight,each of which is a numeric vector.weight gives individual bodyweights in kg, andheight gives individual heights in cm,corresponding to each item in the inputage_months.

Author(s)

Caroline Ring

References

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

Generate serum creatinine values for a virtual population.

Description

Predict serum creatinine from age using smoothing splines and kernel densityestimates of residual variability fitted to NHANES data,, for a givencombination of gender and NHANES race/ethnicity category.

Usage

gen_serum_creatinine(gender, reth, age_years, age_months, nhanes_mec_svy)

Arguments

Details

This function should usually not be called directly by the user. It is usedbyhttkpop_generate() in "virtual-individuals" mode, after drawinggender, NHANES race/ethnicity category, and age from their NHANESproportions/distributions.

Value

A vector of numeric generated serum creatinine values (mg/dL).

Author(s)

Caroline Ring

References

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

Retrieve chemical information on 2023 EPA PFAS Chemicals

Description

This function is a wrapper forget_cheminfo that only listschemicals from the Smeltz, Kreutz, and Crizer data sets collected on PFASbetween 2019 and 2022. Plasma protein binding (fraction unbound) data werecollected using ultracentrifugation (UC) instead of rapid equilibriumdialysis. Intrsinsic hepatic clearance (Clint) data were collected withsubstrate depletion (over time) assays.

Usage

get_2023pfasinfo(  info = "CAS",  species = "Human",  fup.lod.default = 0.005,  model = "3compartmentss",  default.to.human = FALSE,  median.only = FALSE,  fup.ci.cutoff = FALSE,  clint.pvalue.threshold = 0.05,  suppress.messages = FALSE)

Arguments

Details

Note that in some cases theFunbound.plasma and theintrinsic clearance areprovided as a series of numbers separated by commas. These values are the result of Bayesian analysis and characterize a distribution: the first valueis the median of the distribution, while the second and third values are the lower and upper 95th percentile (that is qunatile 2.5 and 97.5) respectively.For intrinsic clearance a fourth value indicating a p-value for a decrease isprovided. Typically 4000 samples were used for the Bayesian analusis, suchthat a p-value of "0" is equivale to "<0.00025". See Wambaugh et al. (2019)for more details. If argument meadian.only == TRUE then only the median isreported for parameters with Bayesian analysis distributions. If the 95credible interval is larger than fup.ci.cutoff (defaultsto NULL) then the Fup is treated as too uncertain and the value NA is given.

Value

Author(s)

John Wambaugh

Examples

PFASCssTable <- NULLfor (this.id in get_2023pfasinfo(info="dtxsid")){  PFASCssTable <- rbind(PFASCssTable, data.frame(    DTXSID = this.id,    Css = try(calc_analytic_css(dtxsid=this.id,                                   model="sumclearancespfas",                                  suppress.messages=TRUE   ))))}

Retrieve in vitro measured Caco-2 membrane permeabilit

Description

This function checks for chemical-specific in vitro measurements of the Caco-2 membrane permeability in thechem.physical_and_invitro.data table. If no value is available argumentCaco2.Pab.default is returned.Anywhere that the values is reported by three numbers separated by a comma (this also happens for plasma protein binding) the three values are: median, lower 95 percent confidence intervals, upper 95 percent confidence interval. Unless you are doing monte carlo work it makes sense to ignore the second and third values.

Usage

get_caco2(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  Caco2.Pab.default = 1.6,  suppress.messages = FALSE)

Arguments

Author(s)

John Wambaugh

Retrieve chemical identity from HTTK package

Description

Given one of chem.name, chem.cas (Chemical Abstract Service Registry Number),or DTXSID (DSStox Substance Identifierhttps://comptox.epa.gov/dashboard) thisfunction checks if the chemical is available and, if so, returns all threepieces of information.

Usage

get_chem_id(chem.cas = NULL, chem.name = NULL, dtxsid = NULL)

Arguments

Value

A list containing the following chemical identifiers:

Author(s)

John Wambaugh and Robert Pearce

Retrieve chemical information available from HTTK package

Description

This function lists information on all the chemicals within HTTK for which there are sufficient data for the specified model and species. By default the function returns only CAS (that is, info="CAS"). The type of information available includes chemical identifiers ("Compound", "CAS", "DTXSID"), in vitromeasurements ("Clint", "Clint.pvalue", "Funbound plasma", "Rblood2plasma"), and physico-chemical information ("Formula", "logMA", "logP", "MW","pKa_Accept", "pKa_Donor"). The argument "info" can be a single type of information, "all" information, or a vector of specific types of information.The argument "model" defaults to "3compartmentss" and the argument "species" defaults to "human". Since different models have different requirements and not all chemicals have complete data, this function will return different numbers of chemicals depending on the model specified. Ifa chemical is not listed by get_cheminfo then either the in vitro orphysico-chemical data needed are currently missing (but could potentiallybe added usingadd_chemtable.

Usage

get_cheminfo(  info = "CAS",  species = "Human",  fup.lod.default = 0.005,  model = "3compartmentss",  default.to.human = FALSE,  median.only = FALSE,  fup.ci.cutoff = TRUE,  clint.pvalue.threshold = 0.05,  physchem.exclude = TRUE,  class.exclude = TRUE,  suppress.messages = FALSE)

Arguments

Details

When default.to.human is set to TRUE, and the species-specific data,Funbound.plasma and Clint, are missing fromchem.physical_and_invitro.data, human values are given instead.

In some cases the rapid equilibrium dialysis method (Waters et al., 2008)fails to yield detectable concentrations for the free fraction of chemical. In those cases we assume the compound is highly bound (that is, Fup approacheszero). For some calculations (for example, steady-state plasma concentration)there is precedent (Rotroff et al., 2010) for using half the average limit of detection, that is, 0.005 (this value is configurable via the argumentfup.lod.default). We do not recommend using other models where quantities like partition coefficients must be predicted using Fup. We alsodo not recommend including the value 0.005 in training sets for Fup predictivemodels.

Note that in some cases theFunbound.plasma (fup) and theintrinsic clearance (clint) areprovided as a series of numbers separated by commas. These values are the result of Bayesian analysis and characterize a distribution: the first valueis the median of the distribution, while the second and third values are the lower and upper 95th percentile (that is quantile 2.5 and 97.5) respectively.For intrinsic clearance a fourth value indicating a p-value for a decrease isprovided. Typically 4000 samples were used for the Bayesian analysis, suchthat a p-value of "0" is equivalent to "<0.00025". See Wambaugh et al. (2019)for more details. If argument median.only == TRUE then only the median isreported for parameters with Bayesian analysis distributions. If the 95credible interval spans the range of 0.1 to 0.9 and fup.ci.cutoff is set to TRUE,i.e., the default setting, then the Fup is treated as too uncertain andthe value NA is given.

Value

Author(s)

John Wambaugh, Robert Pearce, and Sarah E. Davidson

References

Rotroff DM, Wetmore BA, Dix DJ, Ferguson SS, Clewell HJ, Houck KA, LeCluyse EL, Andersen ME, Judson RS, Smith CM, others (2010).“Incorporating human dosimetry and exposure into high-throughput in vitro toxicity screening.”Toxicological Sciences,117(2), 348–358.doi:10.1093/toxsci/kfq220.

Waters NJ, Jones R, Williams G, Sohal B (2008).“Validation of a rapid equilibrium dialysis approach for the measurement of plasma protein binding.”Journal of pharmaceutical sciences,97(10), 4586–4595.doi:10.1002/jps.21317.

Wambaugh JF, Wetmore BA, Ring CL, Nicolas CI, Pearce RG, Honda GS, Dinallo R, Angus D, Gilbert J, Sierra T, others (2019).“Assessing toxicokinetic uncertainty and variability in risk prioritization.”Toxicological Sciences,172(2), 235–251.doi:10.1093/toxsci/kfz205.

Examples

# List all CAS numbers for which the 3compartmentss model can be run in humans: get_cheminfo()get_cheminfo(info=c('compound','funbound.plasma','logP'),model='pbtk') # See all the data for humans:get_cheminfo(info="all")TPO.cas <- c("741-58-2", "333-41-5", "51707-55-2", "30560-19-1", "5598-13-0", "35575-96-3", "142459-58-3", "1634-78-2", "161326-34-7", "133-07-3", "533-74-4", "101-05-3", "330-54-1", "6153-64-6", "15299-99-7", "87-90-1", "42509-80-8", "10265-92-6", "122-14-5", "12427-38-2", "83-79-4", "55-38-9", "2310-17-0", "5234-68-4", "330-55-2", "3337-71-1", "6923-22-4", "23564-05-8", "101-02-0", "140-56-7", "120-71-8", "120-12-7", "123-31-9", "91-53-2", "131807-57-3", "68157-60-8", "5598-15-2", "115-32-2", "298-00-0", "60-51-5", "23031-36-9", "137-26-8", "96-45-7", "16672-87-0", "709-98-8", "149877-41-8", "145701-21-9", "7786-34-7", "54593-83-8", "23422-53-9", "56-38-2", "41198-08-7", "50-65-7", "28434-00-6", "56-72-4", "62-73-7", "6317-18-6", "96182-53-5", "87-86-5", "101-54-2", "121-69-7", "532-27-4", "91-59-8", "105-67-9", "90-04-0", "134-20-3", "599-64-4", "148-24-3", "2416-94-6", "121-79-9", "527-60-6", "99-97-8", "131-55-5", "105-87-3", "136-77-6", "1401-55-4", "1948-33-0", "121-00-6", "92-84-2", "140-66-9", "99-71-8", "150-13-0", "80-46-6", "120-95-6","128-39-2", "2687-25-4", "732-11-6", "5392-40-5", "80-05-7", "135158-54-2", "29232-93-7", "6734-80-1", "98-54-4", "97-53-0", "96-76-4", "118-71-8", "2451-62-9", "150-68-5", "732-26-3", "99-59-2", "59-30-3", "3811-73-2", "101-61-1", "4180-23-8", "101-80-4", "86-50-0", "2687-96-9", "108-46-3", "95-54-5", "101-77-9", "95-80-7", "420-04-2", "60-54-8", "375-95-1", "120-80-9","149-30-4", "135-19-3", "88-58-4", "84-16-2", "6381-77-7", "1478-61-1", "96-70-8", "128-04-1", "25956-17-6", "92-52-4", "1987-50-4", "563-12-2", "298-02-2", "79902-63-9", "27955-94-8")httk.TPO.rat.table <- subset(get_cheminfo(info="all",species="rat"), CAS %in% TPO.cas) httk.TPO.human.table <- subset(get_cheminfo(info="all",species="human"), CAS %in% TPO.cas) # create a data.frame with all the Fup values, we ask for model="schmitt" since# that model only needs fup, we ask for "median.only" because we don't care# about uncertainty intervals here:fup.tab <- get_cheminfo(info="all",median.only=TRUE,model="schmitt")# calculate the median, making sure to convert to numeric values:median(as.numeric(fup.tab$Human.Funbound.plasma),na.rm=TRUE)# calculate the mean:mean(as.numeric(fup.tab$Human.Funbound.plasma),na.rm=TRUE)# count how many non-NA values we have (should be the same as the number of # rows in the table but just in case we ask for non NA values:sum(!is.na(fup.tab$Human.Funbound.plasma))

Retrieve and parse intrinsic hepatic clearance

Description

This function retrieves the chemical- and species-specific intinsichepatic clearance (Cl_int,inits of uL/min/million hepatocytes) fromchem.physical_and_invitro.data. If that parameter is described by a distribution (that is, a median, lower-, upper-95th percentile and p-value separated by commas) this function splits those quantiles into separate values. Most Cl_int values have anaccompanying p-value indicating the probability that no decrease was observed. If the p-values exceeds a threhsold (default 0.05) the clearance isset to zero (no clearance). Some values extracted from the literature do nothave a p-value.

Usage

get_clint(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  species = "Human",  default.to.human = FALSE,  force.human.clint = FALSE,  suppress.messages = FALSE,  clint.pvalue.threshold = 0.05)

Arguments

Value

list containing:

Author(s)

John Wambaugh

See Also

chem.physical_and_invitro.data

Retrieve or calculate fraction of chemical absorbed from the gut

Description

This function checks for chemical-specific in vivo measurements of the fraction absorbed from thegut in thechem.physical_and_invitro.data table. If in vivodata are unavailable (orkeepit100 == TRUE) we attempt to use in vitro Caco-2 membrane permeability to predict the fractions according tocalc_fbio.oral.

Usage

get_fbio(  parameters = NULL,  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  species = "Human",  Caco2.Pab.default = 1.6,  Caco2.Fgut = TRUE,  Caco2.Fabs = TRUE,  overwrite.invivo = FALSE,  keepit100 = FALSE,  suppress.messages = FALSE,  ...)

Arguments

Author(s)

Greg Honda and John Wambaugh

See Also

calc_fbio.oral

Retrieve and parse fraction unbound in plasma

Description

This function retrieves the chemical- and species-specific fractionunbound in plasma (f_up) fromchem.physical_and_invitro.data. If that parameter is described by a distribution (that is, a median, lower-, and upper-95th percentile separated by commas) this function splits those quantiles into separate values.

Usage

get_fup(  chem.cas = NULL,  chem.name = NULL,  dtxsid = NULL,  species = "Human",  default.to.human = FALSE,  force.human.fup = FALSE,  suppress.messages = FALSE,  minimum.Funbound.plasma = 1e-04)

Arguments

Value

list containing:

Author(s)

John Wambaugh

See Also

chem.physical_and_invitro.data

Categorize kidney function by GFR.

Description

For adults: In general GFR > 60 is considered normal 15 < GFR < 60 is considered kidney disease GFR < 15 is considered kidney failure

Usage

get_gfr_category(age_years, age_months, gfr_est)

Arguments

Details

These values can also be used for children 2 years old and greater (see PEDIATRICS INREVIEW Vol. 29 No. 10 October 1, 2008 pp. 335-341 (doi:10.1542/pir.29-10-335))

Value

Vector of GFR categories: 'Normal', 'Kidney Disease', 'KidneyFailure'.

Author(s)

Caroline Ring

References

Ring CL, Pearce RG, Setzer RW, Wetmore BA, Wambaugh JF (2017).“Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability.”Environment International,106, 105–118.doi:10.1016/j.envint.2017.06.004.

Get timeseries containing the change of each of the input parameters.