calcCombCnvProb: Run BADGER combined model to assess posterior probability of...

In JEFworks/badger: Bayesian approach for detecting copy number alterations from single cell RNA-seq data

Description

Run BADGER combined model to assess posterior probability of CNVs given both allele and expression data

Usage

calcCombCnvProb(r, cov.sc, l, cov.bulk, region, gtf, gexp, fits, gos, m,
  filter = TRUE, pe = 0.01, mono = 0.7, n.iter = 1000, quiet = TRUE,
  delim = ":")

Arguments

r	Matrix of alt allele count in single cells
cov.sc	Matrix of coverage in single cells
l	Vector of alt allele count in bulk
cov.bulk	Vector of coverage in bulk
region	Region of interest such as expected CNV boundaries
gtf	GTF file contents for mapping SNPs to genes. Required if plotGene = TRUE
gexp	Normalized gene expression matrix
fits	Fit for variance around mean
gos	Gene position table
m	Expected mean deviation due to copy number change
filter	Boolean for whether to filter out SNP sites with no coverage. Default: TRUE
pe	Effective error rate to capture error from sequencing, etc. Default: 0.01
mono	Rate of mono-allelic expression. Default: 0.7
n.iter	Number of iterations in MCMC. Default: 1000
quiet	Boolean of whether to suppress progress bar. Default: TRUE
delim	Delimiter for names of SNPs as Chromosome[delim]Position. Default: ":" ex. chr1:283838897

Value

List of posterior probabilities for CNV and direction of CNV (deletion vs. amplification)

Examples

data('MM16.counts')
mat <- log2(MM16.counts + 1)
data('Normal.counts')
mat.ref <- log2(Normal.counts + 1)
mats <- normalizedExpression(mat, mat.ref)
gexp <- mats[[1]]
fits <- mvFit(gexp)
region <- data.frame('chr'=1, start=0, end=1e9)
data(snpsHet_MM16ScSample)
data(snpsHet_MM16BulkSample)
## Not run: 
gtfFile <- 'data-raw/Homo_sapiens.GRCh37.75.gtf'
gtf <- read.table(gtfFile, header=F, stringsAsFactors=F, sep='\t')
region <- data.frame('chr'=2, start=0, end=1e9) # deletion region
library(biomaRt) ## for gene coordinates
mart.obj <- useMart(biomart = "ENSEMBL_MART_ENSEMBL", dataset = 'hsapiens_gene_ensembl', host = "jul2015.archive.ensembl.org")
gos <- getBM(values=rownames(mat.tot),attributes=c("ensembl_gene_id","chromosome_name","start_position","end_position"),filters=c("ensembl_gene_id"),mart=mart.obj)
gos$pos <- (gos$start_position + gos$end_position)/2
results <- calcCombCnvProb(r, cov.sc, l, cov.bulk, region, gtf, gexp, fits, gos, m=0.15)

## End(Not run)

JEFworks/badger documentation built on May 7, 2019, 7:40 a.m.