Statistical Sciences & Operations Research Seminar Virginia Commonwealth University

TITLE: A per-base regression model for ChIP-seq peak calling SPEAKER: Dr. Nak-Kyeong Kim, Department of Biostatistics, VCU TIME: 12:00pm-1:00 pm, Thursday, March 3, 2016 PLACE: Room 4119, Grace Harris Hall

ABSTRACT: Chromatin immunoprecipitation followed by high-throughput sequencing (ChIP-seq) can locate transcription factor binding sites on genomic scale. Although many models and programs are available to call peaks, none has dominated its competition in comparison studies. We propose a per-base regression model with a kernel of the normal-exponential two-peak (NEXT-peak) density for calling peaks. The proposed NEXT-peak kernel parallels the physical processes generating the empirical data. The strand-specific, per-base tag count is assumed to be sum of the tag counts from the protein binding and the tag counts from the noise process. Unlike the existing models, the NEXT-peak model estimates strength of binding by computing the mixing probabilities between signal and noise; it also assigns a standard error to an estimated binding location. The comparison study with existing programs on real ChIP-seq datasets demonstrates that the NEXT-peak model performs well both in calling peaks and locating them.

Department of Statistical Sciences and Operations Research www.stat.vcu.edu (804) 828-0001