quantifyExpressionsFromBAMs: Evaluate introns and exons expressions from BAM or SAM files

In ste-depo/INSPEcT: Modeling RNA synthesis, processing and degradation with RNA-seq data

Description

Given a TranscriptDb object and a list of BAM or SAM files "quantifyExpressionsFormBAMs" evaluates exons and introns expressions and the associated variances per each gene.

Usage

quantifyExpressionsFromBAMs(
  txdb,
  BAMfiles,
  experimentalDesign,
  by = c("gene", "tx"),
  countMultiMappingReads = FALSE,
  allowMultiOverlap = FALSE,
  prioritizeExons = TRUE,
  libsize = c("assigned", "all"),
  strandSpecific = 0,
  isPairedEnd = FALSE,
  DESeq2 = TRUE,
  varSamplingCondition = NULL,
  BPPARAM = SerialParam()
)

Arguments

txdb	A TranscriptDB object
BAMfiles	A vector of paths
experimentalDesign	A numerical which reports the desing of the experiment in terms of time points and replicates. Time points must be ordered according to the sequence of files submitted for the analysis, these labels characterize different files as replicates of a given condition.
by	A character, either "gene" or "tx", indicating if expressions and counts should be summarized at the levels of genes or transcripts. "gene" by default. In case "tx" is selected, we suggest to set argument "allowMultiOverlap" to TRUE, otherwise the reads mapping to overlapping transcripts of the same gene will remain unassigned.
countMultiMappingReads	A logical, if multimapping reads should be counted, FALSE by default. Multimap reads are identified using the tag "NH" in the bam/sam file.
allowMultiOverlap	A logical, indicating if a read is allowed to be assigned to more than one feature, FALSE by default
prioritizeExons	A logical, indicating whether reads assigned to exon shold not be accounted for intron counts. If set to FALSE, reads with shared overlap between an exon and the following intron will be assigned also to introns. This could improve intronic quantification in experimental settings (including polyA library preparation) or compact genomes were intronic reads are sampled at a very low rate compared to exonic reads. By default, TRUE.
libsize	A character, either "assigned" or "all", indicating whether the libsize for expression normalization should include all mapped reads or only the reads assigned to any of the features. By default, "assigned" is selected.
strandSpecific	Numeric, 0 if no strand-specific read counting should be performed, 1 stranded, 2 reversely-stranded. 0 by default
isPairedEnd	A logical, if paired-end reads are used, FALSE by default
DESeq2	A logical, if TRUE exons and introns variances are evaluated through the package DESeq2, if FALSE through plgem
varSamplingCondition	A character reporting which experimental condition should be used to sample the variance if DESeq2 = FALSE.
BPPARAM	Parallelization parameters for bplapply. By default SerialParam() By default, the first element of "experimentalDesign" with replicates.

Value

A list containing expressions and associated variances for exons and introns.

Examples

if( Sys.info()["sysname"] != "Windows" ) {
  require(TxDb.Mmusculus.UCSC.mm9.knownGene)
  txdb<-TxDb.Mmusculus.UCSC.mm9.knownGene
  expDes<-c(0,0,1,1)
  
  paths_total<-system.file('extdata/', c('bamRep1.bam'
                                        ,'bamRep2.bam'
                                        ,'bamRep3.bam'
                                        ,'bamRep4.bam')
                          ,package='INSPEcT')
 
  matExp<-quantifyExpressionsFromBAMs(txdb=txdb
                                     ,BAMfiles=paths_total
                                     ,experimentalDesign=expDes)
}

ste-depo/INSPEcT documentation built on Oct. 3, 2020, 9:14 p.m.