Presenter: Yaomin Xu, Ph.D.
Next-gen sequencing techniques, such as ChIP-seq and MBD-seq, are widely used to study epigenetic modifications in biology and disease studies. Genome-wide, high-resolution identification of differential epigenomic events in experiments comparing groups of samples remains an analytical challenge. We propose a novel analysis strategy, iDPT, which integrates signal Decomposition, whole genome Pattern recognition and scan, and differential Testing into a consolidated pipeline, to address this issue. iDPT robustly identified differentially methylated sites in our dataset comparing three groups of normal, non-CIMP and CIMP tumor colon samples. We confirmed the accuracy of iDPT by using the gold standard capillary bisulfite sequencing on a randomly selected subset of identified sites, and obtained a 100% validation rate. In this talk, I will also demonstrate the high reproducibility of iDPT analysis in analyzing two independent datasets. In short, iDPT provides a unified analysis framework for systematic comparison of epigenomic NGS data on groups of clinical or biological samples.
Hosted by: Department of Biostatistics