Nonparametric Inference for ChIP-Seq (NICS)

• Activity Number: 72
• Type: Contributed
• Date/Time: Sunday, July 31, 2011 : 4:00 PM to 5:50 PM
• Sponsor: Biometrics Section
• Author(s): Mingqi Wu*+ and Faming Liang and Monique Rijnkels
• Companies: Merck & Co., Inc. and Texas A & M University and Baylor College of Medicine
• Keywords: ChIP-seq ; Nonparametric inference ; False Discovery Rate

Abstract

Due to its higher resolution mapping and stronger ChIP enrichment signals, ChIP-seq tends to replace ChIP-chip technology in studying genome-wide protein-DNA interactions. While the massive digital ChIP-seq data present new challenges to statisticians. To date, most methods proposed for ChIP-seq data analysis are model based, however finding an appropriate model for each dataset may be difficult, given the complexity of biological systems and the possible bias generated in the sequencing process. In this talk, we present a nonparametric method for the ChIP-seq data analysis, named NICS. The new approach has three attractive features:(i)robustness in dealing with different datasets; (ii)by working on the difference between the averaged treatment and control samples, it avoids the local fluctuations and sample bias; (iii)efficiency in terms of computational cost. In this talk, we also present a simple semi-empirical method for simulating ChIP-seq data, which allows better assessment of the performance of NICS. We compare NICS with some of the existing methods, e.g. MACS, PICS and BayesPeak based on real and simulated datasets. The results indicate that NICS can outperform others.