normLGF {HiTC} | R Documentation |
Parametric model to remove systematic biases in the raw contact maps
normLGF(x, family=c("poisson", "nb"))
x |
object that inherits from class |
family |
parametric model to fit (poisson or nb) |
This function implements the HiCNorm method proposed by Hu et al. Briefly, the method uses a generalized linear model to correct the systematic biases (effective fragment length, GC content, mappability) in a Hi-C contact map.
Returns a HTCexp
object with a normalized contact map.
N. Servant, M. Hu, S. Selvaraj
Hu M, Deng K, Selvaraj S, Qin Z, Ren B, Liu JS. HiCNorm: removing biases in Hi-C data via Poisson regression. Bioinformatics. 2012;28(23):3131-3.
getAnnotatedRestrictionSites
, setGenomicFeatures
## Not run: require(HiTC) require(BSgenome.Hsapiens.UCSC.hg18) ##Lieberman data exDir <- system.file("extdata", package="HiTC") l <- sapply(list.files(exDir, pattern=paste("HIC_gm06690_"), full.names=TRUE), import.my5C) hiC <- HTClist(l) hiC <- hiC[isIntraChrom(hiC)] names(hiC) ## Mappability data From http://hgdownload.cse.ucsc.edu/goldenPath/hg18/encodeDCC/wgEncodeMapability/ map_hg18<- import("wgEncodeCrgMapabilityAlign100mer.bw", format="BigWig") ## Get the genomic feature of the chromosome 12 hiC_annot <- HTClist(lapply(hiC, setGenomicFeatures, resSite="AAGCTT", overhangs5=1, genomePack="BSgenome.Hsapiens.UCSC.hg18", wingc=200, mappability=map_hg18, winmap=500)) hiC_annot$chr12chr12 ## Normalize the data hiCnorm <- HTClist(lapply(hiC_annot, normLGF)) ## End(Not run)