Movatterモバイル変換

Title:

Automated Analysis of Multiplex Digital PCR Data

Version:

2.0.1

Description:

The automated clustering and quantification of the digital PCR data is based on the combination of 'DBSCAN' (Hahsler et al. (2019) <doi:10.18637/jss.v091.i01>) and 'c-means' (Bezdek et al. (1981) <doi:10.1007/978-1-4757-0450-1>) algorithms. The analysis is independent of multiplexing geometry, dPCR system, and input amount. The details about input data and parameters are available in the vignette.

License:

MIT + file LICENSE

Encoding:

UTF-8

Depends:

R (≥ 4.0.0)

Imports:

cluster, dbscan, e1071, exactci, ggplot2, ggpubr, graphics,raster, rlist, scales, shiny, shinyjs, stats, stringr, utils

RoxygenNote:

7.2.1

Suggests:

knitr, rmarkdown, testthat

VignetteBuilder:

knitr

URL:

https://github.com/alfodefalco/dPCP

BugReports:

https://github.com/alfodefalco/dPCP/issues

NeedsCompilation:

Packaged:

2023-08-12 17:53:04 UTC; alfonsodefalco

Author:

Alfonso De Falco

[aut, cre], Michel Mittelbronn [ctb], Christophe Olinger [ctb], Daniel Stieber [ctb], Laboratoire national de santé [cph]

Maintainer:

Alfonso De Falco <alfonsodefalco90@gmail.com>

Repository:

CRAN

Date/Publication:

2023-08-12 18:20:02 UTC

Prediction of clusters centroid position

Description

This function calculates the coodintaes of all clusters centroid.

Usage

centers_data(sample.subquality, sample.table, referenceDB)## S3 method for class 'centers_data'plot(x, ..., sample = "all")

Arguments

sample.subquality

an object of classread_sample, inheritedfromread_sample.

sample.table

object of classsample_table, inherited fromread_sampleTable.

referenceDB

an object of classreference_dbscan, inheritedfromreference_dbscan

x

an object of classcenters_data

...

Arguments to be passed to methods

sample

'all' to show all samples, or a numeric vector indicatingthe row number of samples in the sample table.

Value

An object of classcenters_data containing a sublist foreach sample. Each sublist has the following components:

quality

quality threshold used inread_sample.

reference

reference ID.

centers

a data frame with the centroids coordinates.

data

a data frame with the fluorescence intensities.

Examples

library(dPCP)#Find path of sample table and location of reference and input filessampleTable <- system.file("extdata", "Template_sampleTable.csv",                     package = "dPCP")fileLoc <- system.file("extdata",package = "dPCP")#Read sample table filesample.table <- read_sampleTable(sampleTable, system = "bio-rad",                                 file.location = fileLoc)#Read reference filesref <- read_reference(sample.table, system = "bio-rad",                      file.location = fileLoc)#Read samples filessamp <- read_sample(sample.table, system = "bio-rad",                    file.location = fileLoc)#Reference DBSCAN clusteringdbref <- reference_dbscan(ref, sample.table, save.template = FALSE)#Predict position of clusters centroid from reference DBSCAN resultscent <- centers_data(samp, sample.table,dbref)plot(cent, sample = "all")

Cluster analysis with fuzzy c-means algorithm

Description

This function carries out the c-means cluster analysis, using the centroidsposition as initial values for cluster centers.

Usage

cmeans_clus(centers.data)## S3 method for class 'cmeans_clus'plot(x, ..., sample = "all", color.blind = FALSE)

Arguments

centers.data

an object of classcenters_data, inheritedfromcenters_data.

x

an object of classcmeans_clus

...

Arguments to be passed to methods

sample

'all' to show all samples, or a numeric vector indicatingthe row number of samples in the sample table.

color.blind

logical. If TRUE colors optimized for colorblind readersare used.

Value

An object of classcmeans_clus containing a sublist foreach sample. Each sublist has the following components:

quality

quality threshold used inread_sample.

reference

reference ID.

centers

a data frame with the centroids coordinates.

data

a data frame with the fluorescence intensities and clustersname.

membership

a matrix with the membership values of the data elementsto the clusters. See alsocmeans

Examples

library(dPCP)#Find path of sample table and location of reference and input filessampleTable <- system.file("extdata", "Template_sampleTable.csv",                     package = "dPCP")fileLoc <- system.file("extdata",package = "dPCP")#Read sample table filesample.table <- read_sampleTable(sampleTable, system = "bio-rad",                                 file.location = fileLoc)#Read reference filesref <- read_reference(sample.table, system = "bio-rad",                      file.location = fileLoc)#Read samples filessamp <- read_sample(sample.table, system = "bio-rad",                    file.location = fileLoc)#Reference DBSCAN clusteringdbref <- reference_dbscan(ref, sample.table, save.template = FALSE)#Predict position of clusters centroid from reference DBSCAN resultscent <- centers_data(samp, sample.table,dbref)#Fuzzy c-means clusteringcmclus <- cmeans_clus(cent)plot(cmclus, sample = "all")

Automated analysis of digital PCR data

Description

This function carries out the autometed clustering of digital PCR data.

Usage

dPCP(  file,  system = NULL,  file.location = ".",  reference.quality = 0.5,  sample.quality = 0.5,  eps = 200,  minPts = 50,  save.template = FALSE,  rain = TRUE,  QC.reference = FALSE,  partition.volume = NULL)## S3 method for class 'dPCP'plot(  x,  ...,  sample = "all",  reference = "all",  type = "dPCP",  color.blind = FALSE)

Arguments

file

character. The name or the path of csv file to be read. If itdoes not contain an absolute path, the file name is relative to the currentworking directory, (getwd).

system

character. The name of digital PCR system used to generatethe data. It must be either Thermo Fisher or Bio-Rad. Abbreviations arealso accepted.

file.location

character. Full path name to reference and samplefiles location. The default corresponds to the working directory,(getwd). Tilde expansion (see(path.expand)) is performed.

reference.quality

numeric. Between 0 and 1. Quality threshold tosubset the data. If different thresholds have to be applied to variousreference samples, a vectror of the same length of number of referencesamples has to be provided. Used only when thesystem is ThermoFisher.

sample.quality

numeric. Between 0 and 1. Quality threshold to subsetdata. If different thresholds have to be applied to various samples, avectror of the same length of number of samples has to be provided. Usedonly when thesystem is Thermo Fisher.

eps

numeric. Input parameter for the DBSCAN algorithm.It represents the maximum distance between the elements within a cluster.See alsodbscan. If different values have to beapplied to various reference samples, a vectror of the same length ofnumber of reference samples has to be provided.

minPts

numeric. Input parameter for the DBSCAN algorithm.It represents the number of minimum elements to assemble a cluster. Seealsodbscan. If different values have to beapplied to various reference samples, a vectror of the same length ofnumber of reference samples has to be provided.

save.template

logical. If TRUE a template of DBSCAN analysis ofreference samples is saved. Whensystem is Thermo Fisher,save.template can be also a character vector indicating the chipID.

rain

logical. If TRUE the rain analysis is carried out.

QC.reference

logical. If TRUE the fraction of rain elements in thereference samples is carried out. Warning messages are displayed when thepercentage of rain is high.

partition.volume

numeric. This parameters is taken into account whenthe parameter 'system' is set on Other. Indicate the partion volume inmicroliters spcific to the digital PCR system.

x

an object of classdPCP

...

Arguments to be passed to methods

sample

'all' to show all samples, or a numeric vector indicatingthe row number of samples in the sample table.

reference

'all' to show all reference samples, or a character vectorwith chip ID (Thermo Fisher) or the file name (Bio-rad) of referencesamples to be showed.

type

string. Type of plot to be showed. Available plots:'reference dbscan', 'centers', 'cmeans', 'rain', 'dPCP'.@param color.blind logical. If TRUE colors optimized for colorblindreaders are used.

color.blind

logical. If TRUE colors optimized for colorblind readersare used.

Value

An object of classdPCP containing the following components:

referenceDB

an object of classreference_dbscan.

samples

a list of samples. Each sample sublist contains theinformation about the cluster analysis.

results

an object of classreplicates_quant.

Examples

library(dPCP)#Find path of sample table and location of reference and input filessampleTable <- system.file("extdata", "Template_sampleTable.csv",                     package = "dPCP")fileLoc <- system.file("extdata", package = "dPCP")#dPCP analysisresults <- dPCP(sampleTable, system = "bio-rad", file.location = fileLoc,                eps = 200, minPts = 50, save.template = FALSE, rain = TRUE,                QC.reference = FALSE)plot(results, sample = 1, type = "dPCP")

Test eps and minPts combinations for DBSCAN analysis

Description

This function tests all combinations of eps and minPts for DBSCAN analysisof reference samples indicated in refID. The results are represented inscatterplots exported to a pdf file.

Usage

dbscan_combination(  refID,  system = NULL,  file.location = ".",  reference.quality = 0.5,  eps = c(120, 150, 180, 200),  minPts = c(20, 50, 80, 100))

Arguments

refID

a string or a character vector of chipID (Thermo Fisher) orthe complete file name with the extension (Bio-Rad) of reference sample(s)to be analysed.

system

character. The name of digital PCR system used to generatethe data. It must be either Thermo Fisher or Bio-Rad. Abbreviations arealso accepted.

file.location

character. Full path name to reference and samplefiles location. The default corresponds to the working directory,(getwd). Tilde expansion (see(path.expand)) is performed.

reference.quality

numeric. Between 0 and 1. Quality thresholdto subset the data (just for Thermo Fisher). If different thresholds haveto be applied to various reference samples, a vectror of the same lengthofrefID has to be provided.

eps

a numeric vector of values to be tested. Maximum distancebetween elements within a cluster in a DBSCAN analysis.See alsodbscan.

minPts

a numeric vector of values to be tested. Number of minimumelements to assemble a cluster in a DBSCAN analysis.See alsodbscan.

Value

A pdf file containing the scatterplots of DBSCAN analysis performedwith all combinations of eps and minPts.Each reference generates a different pdf file.

Examples

library(dPCP)#Find path of sample table and location of reference and input filessampleTable <- system.file("extdata", "Template_sampleTable.csv",                     package = "dPCP")fileLoc <- system.file("extdata", package = "dPCP")dbscan_combination("dilution20200313_B01_Amplitude.csv",                   file.location = fileLoc, system = "bio-rad",                   eps = c(150, 160, 180, 190), minPts = c(80, 100, 120))unlink("dilution20200313_B01_Amplitude.pdf")

Export dPCP analysis results to a csv file

Description

This function exports dPCP analysis results to a csv file.

Usage

export_csv(data, filename)

Arguments

data

an object of classdPCP,target_quant orreplicates_quant.

filename

character. File name (no extension) for csv and pdf files tocreate on disk.

Value

A csv file with the information and results of dPCP analysis.

Examples

library(dPCP)#Find path of sample table and location of reference and input filessampleTable <- system.file("extdata", "Template_sampleTable.csv",                     package = "dPCP")fileLoc <- system.file("extdata",package = "dPCP")#dPCP analysisresults <- dPCP(sampleTable, system = "bio-rad", file.location = fileLoc,                eps = 200, minPts = 50, save.template = FALSE,                rain = TRUE)export_csv(results, filename = "dPCRproject_1")

Manual correction of dPCP cluster analysis

Description

This function builds an interactive app to manually correct the dPCPcluster analysis.

Usage

manual_correction(  data,  filename,  save.plot = FALSE,  format = "png",  dpi = 300,  color.blind = FALSE)

Arguments

data

an object of classdPCP, inherited fromdPCP.

filename

character. File name (no extension) for csv and pdf files tocreate on disk.

save.plot

logical. If TRUE the plots are exported to a file.

format

a string indicating the file format for the export.Available formats: 'eps', 'ps', 'tex', 'pdf', 'jpeg', 'tiff', 'png','bmp', 'svg', 'wmf'.

dpi

numeric. Image resolution.

color.blind

logical. If TRUE colors optimized for colorblindreaders are used.

Value

A Shiny session.

Examples

library(dPCP)#Find path of sample table and location of reference and input filessampleTable <- system.file("extdata", "Template_sampleTable.csv",                     package = "dPCP")fileLoc <- system.file("extdata",package = "dPCP")#dPCP analysisresults <- dPCP(sampleTable, system = "bio-rad", file.location = fileLoc,                eps = 200, minPts = 50, save.template = FALSE,                rain = TRUE)manual_correction(results, filename = "manual_dPCR", save.plot = FALSE)

Identification and clustering of "rain" data

Description

This function identifies the "rain" elements and re-clusters them using theMahalanobis distance. Each "rain" element is assigned to the cluster whoseMahalanobis distance is the lowest.

Usage

rain_reclus(cmeans.cluster)## S3 method for class 'rain_reclus'plot(x, ..., sample = "all", color.blind = FALSE)

Arguments

cmeans.cluster

an object of classcmeans_clus, inheritedfromcmeans_clus.

x

an object of classrain_reclus

...

Arguments to be passed to methods

sample

'all' to show all samples, or a numeric vector indicatingthe row number of samples in the sample table.

color.blind

logical. If TRUE colors optimized for colorblind readersare used.

Value

An object of classrain_reclus containing a sublist foreach sample. Each sublist has the following components:

quality

quality threshold used inread_sample.

reference

reference ID.

centers

a data frame with the centroids coordinates.

data

a data frame with the fluorescence intensities and clustersname.

Examples

library(dPCP)#Find path of sample table and location of reference and input filessampleTable <- system.file("extdata", "Template_sampleTable.csv",                     package = "dPCP")fileLoc <- system.file("extdata",package = "dPCP")#Read sample table filesample.table <- read_sampleTable(sampleTable, system = "bio-rad",                                 file.location = fileLoc)#Read reference filesref <- read_reference(sample.table, system = "bio-rad",                      file.location = fileLoc)#Read samples filessamp <- read_sample(sample.table, system = "bio-rad",                   file.location = fileLoc)#Reference DBSCAN clusteringdbref <- reference_dbscan(ref, sample.table, save.template = FALSE)#Predict position of clusters centroid from reference DBSCAN resultscent <- centers_data(samp, sample.table,dbref)#Fuzzy c-means clusteringcmclus <- cmeans_clus(cent)#Rain classification.rainclus <- rain_reclus(cmclus)plot(rainclus, sample = "all")

Read reference files

Description

This function reads the results files of reference samples listed in thesample table. Fluoresce intensity and quality value (just for Thermo Fisher)are collected.If areference_dbscan template file with the same inputparamters (reference ID, eps, minPts) is available, fluorescence data,quality value and dbscan analysis results are retrived from the templatefile.

Usage

read_reference(  sample.table,  system = NULL,  file.location = ".",  reference.quality = 0.5,  eps = NULL,  minPts = NULL)

Arguments

sample.table

object of classsample_table, inherited fromread_sampleTable.

system

character. The name of digital PCR system used to generatethe data. It must be either Thermo Fisher or Bio-Rad. Abbreviations arealso accepted.

file.location

character. Full path name to reference and samplefiles location. The default corresponds to the working directory,(getwd). Tilde expansion (see(path.expand)) is performed.

reference.quality

eps,minPts

numeric. Input parameters for the DBSCAN algorithm. Ifthey match the paramters ofreference_dbscan template file,the data are retrived from the template.

Value

An object of classread_reference containing a sublist foreach reference. Each sublist has the following components:

quality

value of thereference.quality parameter.

data

a matrix with the fluorescence intensities and qualityvalues.

dbscan

an object of classdbscan_fast, inherited fromdbscan. This component is available only if areference_dbscan template file is used to retrive thedata.

Examples

library(dPCP)#Find path of sample table and location of reference and input filessampleTable <- system.file("extdata", "Template_sampleTable.csv",                     package = "dPCP")fileLoc <- system.file("extdata", package = "dPCP")#Read sample table filesample.table <- read_sampleTable(sampleTable, system = "bio-rad",                                 file.location = fileLoc)#Read reference filesref <- read_reference(sample.table, system = "bio-rad",                      file.location = fileLoc)

Read sample files

Description

This function reads the results files of samples listed in the sample table.Fluoresce intensity and quality value (just for Thermo Fisher) arecollected.

Usage

read_sample(  sample.table,  system = NULL,  file.location = ".",  sample.quality = 0.5,  partition.volume = NULL)

Arguments

sample.table

object of classsample_table, inherited fromread_sampleTable.

system

character. The name of digital PCR system used to generatethe data. It must be either Thermo Fisher or Bio-Rad. Abbreviations arealso accepted.

file.location

character. Full path name to reference and samplefiles location. The default corresponds to the working directory,(getwd). Tilde expansion (see(path.expand)) is performed.

sample.quality

partition.volume

numeric. This parameters is taken into account whenthe parameter 'system' is set on Other. Indicate the partion volume inmicroliters spcific to the digital PCR system.

Value

An object of classread_sample containing a sublist for eachsample. Each sublist has the following components:

quality

value of thesample.quality parameter.

data

a matrix with the fluorescence intensities and qualityvalues.

Examples

library(dPCP)#Find path of sample table and location of reference and input filessampleTable <- system.file("extdata", "Template_sampleTable.csv",                     package = "dPCP")fileLoc <- system.file("extdata", package = "dPCP")#Read sample table filesample.table <- read_sampleTable(sampleTable, system = "bio-rad",                                 file.location = fileLoc)#Read reference filesref <- read_reference(sample.table, system = "bio-rad",                      file.location = fileLoc)#Read samples filessamp <- read_sample(sample.table, system = "bio-rad",                    file.location = fileLoc)

Read sample table

Description

This function reads a file containing the essential information about thesamples and experimental settings. The file has to be filled out by the userand formatted as described in the vignette.

Usage

read_sampleTable(file, system = NULL, file.location = ".")

Arguments

file

character. The name or the path of csv file to be read. If itdoes not contain an absolute path, the file name is relative to the currentworking directory, (getwd).

system

character. The name of digital PCR system used to generatethe data. It must be either Thermo Fisher or Bio-Rad. Abbreviations arealso accepted.

file.location

character. Full path name to reference and samplefiles location. The default corresponds to the working directory,(getwd). Tilde expansion (see(path.expand)) is performed.

Value

An object of classsample_table.

Examples

library(dPCP)#Find path of sample table and location of reference and input filessampleTable <- system.file("extdata", "Template_sampleTable.csv",                     package = "dPCP")fileLoc <- system.file("extdata", package = "dPCP")#Read sample table filesample.table <- read_sampleTable(sampleTable, system = "bio-rad",                                 file.location = fileLoc)

Find the empty partitions and single target clusters in the reference sample

Description

This function computes a DBSCAN analysis to identify single target clustersin the reference samples listed in the sample table.If areference_dbscan template file with the same inputparamters (reference ID, eps, minPts) is available, data are retrivedfrom the template file.

Usage

reference_dbscan(  reference.subquality,  sample.table,  eps = 200,  minPts = 50,  save.template = FALSE)## S3 method for class 'reference_dbscan'plot(x, ..., reference = "all")

Arguments

reference.subquality

an object of classread_reference,inherited fromread_reference.

sample.table

object of classsample_table, inherited fromread_sampleTable.

eps,minPts

numeric. Input parameters for the DBSCAN algorithm. Ifthey match the paramters ofreference_dbscan template file,the data are retrived from the template.

save.template

logical. If TRUE a template of DBSCAN analysis ofreference samples is saved. Whensystem is Thermo Fisher,save.template can be also a character vector indicating the chipID.

x

an object of classreference_dbscan

...

Arguments to be passed to methods

reference

'all' to show all reference samples, or a character vectorwith chip ID (Thermo Fisher) or the file name (Bio-rad) of referencesamples to be showed.

Value

An object of classreference_dbscan containing a sublist foreach reference. Each sublist has the following components:

quality

quality threshold used inread_reference.

data

a matrix with the fluorescence intensities and qualityvalues.

dbscan

an object of classdbscan_fast, inherited fromdbscan.

Examples

library(dPCP)#Find path of sample table and location of reference and input filessampleTable <- system.file("extdata", "Template_sampleTable.csv",                     package = "dPCP")fileLoc <- system.file("extdata",package = "dPCP")#Read sample table filesample.table <- read_sampleTable(sampleTable, system = "bio-rad",                                 file.location = fileLoc)#Read reference filesref <- read_reference(sample.table, system = "bio-rad",                      file.location = fileLoc)#Read samples filessamp <- read_sample(sample.table, system = "bio-rad",                    file.location = fileLoc)#Reference DBSCAN clusteringdbref <- reference_dbscan(ref, sample.table, save.template = FALSE)plot(dbref, reference = "all")

Calculation of targets concentration, pooling the sample replicates

Description

This function calculates the concentration of the targets, combining theresults of the replicates of each sample.

Usage

replicates_quant(raw.results, sample.table)

Arguments

raw.results

an object of classtarget_quant, inheritedfromtarget_quant.

sample.table

object of classsample_table, inherited fromread_sampleTable.

Value

An object of classreplicates_quant containing a sublist forevery sample. Each sublist has the following components:

quality

quality threshold used inread_sample.

reference

reference ID.

raw results

a data frame with the results of quantification.

replicates results

a data frame with the results of quantificationof pooled replicates.

Examples

library(dPCP)#Find path of sample table and location of reference and input filessampleTable <- system.file("extdata", "Template_sampleTable.csv",                     package = "dPCP")fileLoc <- system.file("extdata", package = "dPCP")#Read sample table filesample.table <- read_sampleTable(sampleTable, system = "bio-rad",                                 file.location = fileLoc)#Read reference filesref <- read_reference(sample.table, system = "bio-rad",                      file.location = fileLoc)#Read samples filessamp <- read_sample(sample.table, system = "bio-rad",                   file.location = fileLoc)#Reference DBSCAN clusteringdbref <- reference_dbscan(ref, sample.table, save.template = FALSE)#Predict position of clusters centroid from reference DBSCAN resultscent <- centers_data(samp, sample.table,dbref)#Fuzzy c-means clusteringcmclus <- cmeans_clus(cent)#Rain classification.rainclus <- rain_reclus(cmclus)#Quantificationquantcm <- target_quant(cmclus, sample.table)quant <- target_quant(rainclus, sample.table)#Replicates poolingrep.quant <- replicates_quant(quant, sample.table)

Export dPCP analysis results to a pdf report

Description

This function generates a pdf report of the dPCP analysis.

Usage

report_dPCP(data, filename, sample = "all", color.blind = FALSE)

Arguments

data

an object of classdPCP, inherited fromdPCP.

filename

character. File name (no extension) for csv and pdf files tocreate on disk.

sample

'all' to show all samples, or a numeric vector indicatingthe row number of samples in the sample table.

color.blind

logical. If TRUE colors optimized for colorblindreaders are used.

Value

A pdf file with the information and results of the dPCP analysis.

Examples

library(dPCP)#Find path of sample table and location of reference and input filessampleTable <- system.file("extdata", "Template_sampleTable.csv",                     package = "dPCP")fileLoc <- system.file("extdata", package = "dPCP")#dPCP analysisresults <- dPCP(sampleTable, system = "bio-rad", file.location = fileLoc,                eps = 200, minPts = 50, save.template = FALSE,                rain = TRUE)report_dPCP(results, filename = "dPCRproject_1")

Calculation of targets concentration.

Description

This function calculates the concentration of the targets according to thePoisson distribution.

Usage

target_quant(data.cluster, sample.table)

Arguments

data.cluster

an object of classrain_reclus orcmeans_clus.

sample.table

object of classsample_table, inherited fromread_sampleTable.

Value

An object of classtarget_quant containing a sublist foreach sample. Each sublist has the following components:

quality

quality threshold used inread_sample.

reference

reference ID.

raw results

a data frame with the results of the quantification.

Examples

library(dPCP)#Find path of sample table and location of reference and input filessampleTable <- system.file("extdata", "Template_sampleTable.csv",                     package = "dPCP")fileLoc <- system.file("extdata", package = "dPCP")#Read sample table filesample.table <- read_sampleTable(sampleTable, system = "bio-rad",                                 file.location = fileLoc)#Read reference filesref <- read_reference(sample.table, system = "bio-rad",                      file.location = fileLoc)#Read samples filessamp <- read_sample(sample.table, system = "bio-rad",                    file.location = fileLoc)#Reference DBSCAN clusteringdbref <- reference_dbscan(ref, sample.table, save.template = FALSE)#Predict position of clusters centroid from reference DBSCAN resultscent <- centers_data(samp, sample.table,dbref)#Fuzzy c-means clusteringcmclus <- cmeans_clus(cent)#Rain classification.rainclus <- rain_reclus(cmclus)#Quantificationquantcm <- target_quant(cmclus, sample.table)quant <- target_quant(rainclus, sample.table)