Finding patterns in high-throughput sequencing data

Abstract
“DNA Encodes RNA, RNA Encodes Protein.” Even 60 years after its postulation, the simple, yet exceptionally powerful central dogma of molecular biology still forms the backbone of our understanding of living cells. However, with the on-going revolution in biology we are not only opening a unique window, shedding new light on the instruction book, but we also find ourselves confronted with an increasing number of more, and evermore complex questions. The process of gene expression, from DNA to protein, is controlled in a myriad of ways, each contributing to the huge observable diversity and the dynamics with which nature expresses itself. To probe the individual regulatory mechanisms, high-throughput sequencing technologies have been creatively combined with other biochemical assays: for instance, we can now measure genomic sequences (DNA-Seq), DNA methylation (BS-Seq), RNA abundances (RNA-Seq), protein binding to DNA (ChIP-Seq) or to RNAs (CLIP-Seq or CRAC-Seq), and many other biologically relevant processes. Statistical analysis of these data sets, however, remains challenging. In this talk I will explore how machine-learning techniques can be employed to find hidden patterns in high-throughput sequencing data.