Models and algorithms for detecting DNA copy number variation using next generation sequencing

Friday, November 08, 2013 — Poster Session IV
2:00 p.m. – 4:00 p.m.	FAES Academic Center (Upper-Level Terrace)	NHGRI	COMPBIO-24

Authors

k Ying
ZD Wang
N Hansen
R Shen
J.C Mullikin

Abstract

Structural variation and especially copy number variation (CNV), play an important role in many complex diseases such as autism, Alzheimer’s disease and cancer. Recent advances in high-throughput DNA sequencing technologies such as whole genome or whole exome shotgun have enabled single nucleotide and short deletion/insertion variant detection across many samples with reasonable cost. However, current analytical methods for CNV detection areis still unsatisfactory both in sensitivity and accuracy. Our method combines several features, such as read depth, allele frequency aberration, paired end distributions and split reads in a unified statistical model at each base pair applying a Hidden Markov Model to detect contiguous genomic regions that show CNV signals. Our method can also consider family pedigree information to increase detection power based on Mendelian inheritance laws. Both simulated and real data (family trio sequence from the 1000 Genomes Project) show that our method has significant advantage compared to other methods.