calcAlleleCnvBoundaries: Use HMM to identify potential CNV boundaries based on...

In JEFworks/HoneyBADGER: HMM-integrated Bayesian approach for detecting CNV and LOH events using single-cell RNA-seq data

Description

Use HMM to identify potential CNV boundaries based on patterns of persistent allelic imbalance

Usage

calcAlleleCnvBoundaries(r.maf, n.sc, l.maf, n.bulk, snps, geneFactor,
  min.traverse = 3, t = 1e-06, pd = 0.1, pn = 0.45, min.num.snps = 5,
  trim = 0.1, verbose = FALSE, ...)

Arguments

r.maf	Matrix of alt allele count in single cells.
n.sc	Matrix of site coverage count in single cells.
l.maf	Vector of alt allele count in pooled single cells or bulk.
n.bulk	Vector of site coverage count in pooled single cells or bulk.
snps	SNP annotations
geneFactor	Output of setGeneFactors
min.traverse	Depth traversal to look for subclonal CNVs. Higher depth, potentially smaller subclones detectable. (default: 3)
t	HMM transition parameter. Higher number, more transitions. (default: 1e-6)
pd	Probability of lesser allele detection in deleted region (ie. due to error)
pn	Probability of lesser allele detection in neutral region (ie. 0.5 - error rate)
min.num.snps	Minimum number of snps in candidate CNV
trim	Trim boundary SNPs
verbose	Verbosity(default: FALSE)
...	Additional parameters to pass to calcAlleleCnvProb

Examples

{
data(r)
data(cov.sc)
allele.mats <- setAlleleMats(r, cov.sc)
library(TxDb.Hsapiens.UCSC.hg19.knownGene)
geneFactor <- setGeneFactors(allele.mats$snps, TxDb.Hsapiens.UCSC.hg19.knownGene)
potentialCnvs <- calcAlleleCnvBoundaries(allele.mats$r.maf, allele.mats$n.sc, 
    allele.mats$l.maf, allele.mats$n.bulk, allele.mats$snps, geneFactor)
## visualize affected regions
plotAlleleProfile(allele.mats$r.maf, allele.mats$n.sc, allele.mats$l.maf, 
    allele.mats$n.bulk, allele.mats$snps, region=potentialCnvs$region)
}

JEFworks/HoneyBADGER documentation built on July 24, 2021, 3:01 p.m.