Movatterモバイル変換

[0]ホーム
URL:

Encoding:UTF-8
Type:Package
Version:1.0.3
Date:2019-05-15
Title:Bayesian Multivariate Analysis of Summary Statistics
Description:Multivariate tool for analyzing genome-wide association study results in the form of univariate summary statistics. The goal of 'bmass' is to comprehensively test all possible multivariate models given the phenotypes and datasets provided. Multivariate models are determined by assigning each phenotype to being either Unassociated (U), Directly associated (D) or Indirectly associated (I) with the genetic variant of interest. Test results for each model are presented in the form of Bayes factors, thereby allowing direct comparisons between models. The underlying framework implemented here is based on the modeling developed in "A Unified Framework for Association Analysis with Multiple Related Phenotypes", M. Stephens (2013) <doi:10.1371/journal.pone.0065245>.
License:GPL (≥ 3)
URL:https://github.com/mturchin20/bmass
BugReports:https://github.com/mturchin20/bmass/issues
Depends:R (≥ 3.3.0)
Imports:utils, stats
Suggests:testthat, knitr, rmarkdown
LazyData:true
NeedsCompilation:no
RoxygenNote:6.1.1
VignetteBuilder:knitr
Packaged:2019-05-16 16:23:07 UTC; mturchin20
Author:Michael Turchin [aut, cre], Matthew Stephens [aut], Peter Carbonetto [ctb]
Maintainer:Michael Turchin <mturchin20@uchicago.edu>
Repository:CRAN
Date/Publication:2019-05-17 07:20:12 UTC

Get Marginal {U,D,I} Posteriors

Description

Get marginal posteriors for how much every individualphenotype belongs to categories {U,D,I} across each SNP

Usage

GetMarginalPosteriors(DataSources, ListSNPs, Models, LogFile)

Arguments

DataSources

A string indicating the variable names of theinput datafiles and phenotypes.

ListSNPs

A list produced from runningbmass containing the SNPs of interest to get marginal posteriors for.

Models

A matrix describing the models being explored(default output from runningbmass).

LogFile

A matrix of string outputs for function loggingpurposes (default output from runningbmass).

Value

A list containing three matrices of SNPs x Phenotypesmarginal posteriors for each category {U,D,I}; this list is appended to the input ListSNPs as a new object,Marginals (the full returned object is a list containing the input ListSNPs and the input LogFile).

Examples

Phenotypes <- c("bmass_SimulatedData1", "bmass_SimulatedData2")bmassOutput <- bmass(Phenotypes, bmass_SimulatedSigSNPs)bmassOutput[c("PreviousSNPs", "LogFile")] <-GetMarginalPosteriors(Phenotypes, bmassOutput$PreviousSNPs,bmassOutput$Models, bmassOutput$LogFile)bmassOutput$PreviousSNPs$Marginals

Get Model Prior Matrix

Description

Creates a matrix containing the model descriptions andtheir associated priors.

Usage

GetModelPriorMatrix(DataSources, Models, ModelPriors, LogFile,  SigmaAlphas = c(0.005, 0.0075, 0.01, 0.015, 0.02, 0.03, 0.04, 0.05,  0.06, 0.07, 0.08, 0.09, 0.1, 0.15))

Arguments

DataSources

A string indicating the variable names of theinput datafiles and phenotypes.

Models

A matrix describing the models being explored(default output from runningbmass).

ModelPriors

A vector containing the priors on each modelacross each tranche of sigma alpha (default output from runningbmass; length is number of models times number of sigma alphas).

LogFile

A matrix of string outputs for function loggingpurposes (default output from runningbmass).

SigmaAlphas

A vector containing the different valuestraversed for this 'effect size controlling' hyperparameter (see"Prior on Sigma_Alpha" in Stephens 2013 PLoS ONE,https://doi.org/10.1371/journal.pone.0065245).

Value

A matrix containing the original description of each modelsort by prior, each model's trained prior, the cummulative priordistribution, and the model's original order position.

Examples

Phenotypes <- c("bmass_SimulatedData1", "bmass_SimulatedData2")bmassOutput <- bmass(Phenotypes,bmass_SimulatedSigSNPs)bmassOutput[c("ModelPriorMatrix", "LogFile")] <-   GetModelPriorMatrix(Phenotypes, bmassOutput$Models,  bmassOutput$ModelPriors, bmassOutput$LogFile)head(bmassOutput$ModelPriorMatrix)

Get Top Multivariate Models

Description

Get a summary of the top models per SNP across allmultivariate {U,D,I} combinations based on posterior probabilities.

Usage

GetTopModelsPerSNPViaPosteriors(DataSources, ListSNPs, ModelPriorMatrix,  LogFile)

Arguments

DataSources

A string indicating the variable names of theinput datafiles and phenotypes.

ListSNPs

A list produced from runningbmass containing the SNPs of interest to get marginal posteriors for.

ModelPriorMatrix

A matrix detailing the models beingexplored and their associated priors (obtained by runningGetModelPriorMatrix)

LogFile

A matrix of string outputs for function loggingpurposes (default output from runningbmass).

Value

A matrix containing each model that was a SNP's top modelat least once, along with related information; this matrix isappended to the input ListSNPs as a new object,TopModels (the full returned object is a list containing the input ListSNPs andthe input LogFile).

Examples

Phenotypes <- c("bmass_SimulatedData1", "bmass_SimulatedData2")bmassOutput <- bmass(Phenotypes, bmass_SimulatedSigSNPs)bmassOutput[c("ModelPriorMatrix", "LogFile")] <-   GetModelPriorMatrix(Phenotypes, bmassOutput$Models,  bmassOutput$ModelPriors, bmassOutput$LogFile)bmassOutput[c("PreviousSNPs", "LogFile")] <-  GetTopModelsPerSNPViaPosteriors(Phenotypes,  bmassOutput$PreviousSNPs, bmassOutput$ModelPriorMatrix, bmassOutput$LogFile)head(bmassOutput$PreviousSNPs$TopModels)

GlobalLipids2013 GWAS SNPs

Description

A list of the univariate GWAS significant SNPs fromthe GlobalLipids2013 dataset to be used in the second introductory bmass vignette.

Format

A data frame with 157 rows and 2 variables:

Chr

chromosome

BP

basepair position

Source

Supplementary Tables 2 and 3 fromhttps://doi.org/10.1038/ng.2797.

Bayesian multivariate analysis of summary statistics (bmass)

Description

Runbmass on a set of phenotypes that each haveunivariate GWAS statistics on the same set of SNPs

Usage

bmass(DataSources, GWASsnps = NULL,  SNPMarginalUnivariateThreshold = 1e-06,  SNPMarginalMultivariateThreshold = 1e-06, GWASThreshFlag = TRUE,  GWASThreshValue = 5e-08, NminThreshold = 0,  PrintMergedData = FALSE, PrintProgress = FALSE, ...)

Arguments

DataSources

A string indicating the variable names of theinput datafiles and phenotypes. No default value.

GWASsnps

A data.table containing rows of SNPs that wereunivariate genome-wide significant in the phenotypes being used foranalysis;GWASsnps input file should have two columns, one for chromosome and another for basepair position (with column headersofChr andBP). No default value.

SNPMarginalUnivariateThreshold

A numerical value indicatingthe univariate p-value threshold to use when collecting marginallysignificant SNPs for finalbmass analysis. Default is1e-6.

SNPMarginalMultivariateThreshold

A numerical valueindicating the basic multivariate p-value threshold to use whencollecting marginally significant SNPs for finalbmass analysis. Default is1e-6.

GWASThreshFlag

A logicalTRUE/FALSE flag that indicates whether to threshold inputGWASsnps list by a univariate GWAS p-value or not (eg the inputGWASsnps list contains variants that are significant from discovery + replication data, but the input summary statistics are just from the discovery cohort). Default isTRUE.

GWASThreshValue

A numerical value indicating the univariatep-value threshold to use in conjunction with theGWASThreshFlag.Default is5e-8.

NminThreshold

A numerical value that indicates a sample sizethreshold to use where SNPs below which are removed. Default is0.

PrintMergedData

A logicalTRUE/FALSE flag that indicates whether the intermediary 'merged datafile' should be included in the finalbmass output; this file combines all the phenotypes for every SNP provided just prior to thresholding for marginallysignificant SNPs. Default isFALSE.

PrintProgress

A logicalTRUE/FALSE flag that indicates whether progress statements should be printed tostderr during the course of runningbmass or not.Default isFALSE.

...

Additional optional arguments.

Value

A list containing model, SNP, and posterior information forboth the previously significant univariate SNPs (PreviousSNPs)and the newly significant multivariate SNPs (NewSNPs). For a full breakdown of thebmass output list structure, please see the associated vignettes.

Other Examples

bmass(c("HDL","LDL","TG","TC"), GWASsnps, NminThreshold = 50000)bmass(c("HDL","LDL","TG","TC"), GWASsnps, GWASThreshValue = 1e-8, NminThreshold = 50000, PrintProgress = TRUE)bmass(c("HDL", "LDL", "TG", "TC"), GWASsnps, GWASThreshFlag = FALSE, SNPMarginalUnivariateThreshold = 1e-4, SNPMarginalMultivariateThreshold = 1e-4, PrintMergedData = TRUE)bmassOutput <- bmass(c("HDL","LDL","TG","TC"), GWASsnps, NminThreshold = 50000)

Examples

Phenotypes <- c("bmass_SimulatedData1", "bmass_SimulatedData2")bmassOutput <- bmass(Phenotypes, bmass_SimulatedSigSNPs)summary(bmassOutput)bmassOutput$NewSNPs$SNPs

bmass Simulated Dataset 1

Description

A manually created sample dataset for use in Roxygen2documents and vignettes.

Format

A data frame with 11 rows and 9 variables:

Chr

chromosome

BP

basepair position

Marker

rsID# or other identifier

MAF

Minor Allele Frequency

A1

reference allele

A2

alternative allele

Direction

direction of associationeffect size, + or -

pValue

p-Value of GWAS association

N

sample size

Source

Manually created

bmass Simulated Dataset 2

Description

A manually created sample dataset for use in Roxygen2documents and vignettes.

Format

A data frame with 11 rows and 9 variables:

Chr

chromosome

BP

basepair position

Marker

rsID# or other identifier

MAF

Minor Allele Frequency

A1

reference allele

A2

alternative allele

Direction

direction of associationeffect size, + or -

pValue

p-Value of GWAS association

N

sample size

Source

Manually created

bmass Simulated GWAS SNPs

Description

A manually created list of GWAS significant SNPs to beused in conjunction with 'bmass_SimulatedData1' and 'bmass_SimulatedData2'.

Format

A data frame with 2 rows and 2 variables:

Chr

chromosome

BP

basepair position

Source

Manually created

bmass Test Dataset 1

Description

A manually created sample dataset for use in unittests.

Format

A data frame with 11 rows and 9 variables:

Chr

chromosome

BP

basepair position

Marker

rsID# or other identifier

MAF

Minor Allele Frequency

A1

reference allele

A2

alternative allele

Direction

direction of associationeffect size, + or -

pValue

p-Value of GWAS association

N

sample size

Source

Manually created

bmass Test Dataset 2

Description

A manually created sample dataset for use in unittests.

Format

A data frame with 11 rows and 9 variables:

Chr

chromosome

BP

basepair position

Marker

rsID# or other identifier

MAF

Minor Allele Frequency

A1

reference allele

A2

alternative allele

Direction

direction of associationeffect size, + or -

pValue

p-Value of GWAS association

N

sample size

Source

Manually created

bmass Test GWAS SNPs

Description

A manually created list of GWAS significant SNPs to beused in conjunction with 'bmass_TestData1' and 'bmass_TestData2'.

Format

A data frame with 2 rows and 2 variables:

Chr

chromosome

BP

basepair position

Source

Manually created

[8]ページ先頭
©2009-2025 Movatter.jp