HaplotypeCN: Copy Number Haplotype Inference with Hidden Markov Model and Localized Haplotype Clustering

摘要: Copy number variation (CNV) has been reported to be associated with disease and various cancers. Hence, identifying the accurate position type of CNV is currently a critical issue. There are many tools targeting on detecting regions, constructing haplotype phases or estimating numerical copy numbers. However, none them can do all three tasks at same time. This paper presents method based Hidden Markov Model detect parent specific change both chromosomes signals from SNP arrays. A tree constructed dynamic branch merging model transition status two alleles assessed each locus. The emission models for genotypes formed haplotypes. proposed provide segmentation points regions as well phasing allelic chromosome. estimated numbers provided fractional numbers, which accommodate somatic mutation in cancer specimens that usually consist heterogeneous cell populations. algorithm evaluated simulated data previously published 270 HapMap individuals. results were compared five popular methods: PennCNV, genoCN, COKGEN, QuantiSNP cnvHap. application oral samples demonstrates how facilitate clinical association studies. exhibits comparable sensitivity best our genome-wide study highest detection rate dense regions. In addition, we better accuracy than similar approaches. carried out estimate provides power integer states.