Jordan Eizenga, PhD Student, Biomolecular Engineering & Bioinformatics
Tuesday, May 1, 2018 – 9:00am
Location – Biomedical Sciences, Room 200
Host – Assistant Professor Benedict Paten
Abstract: Much of current computational genomics is built around reference genomes. However, the current paradigm for reference genomes in limited in that it uses a single sequence to represent the genomes of an entire species. Recent research efforts have generalized this concept to graph-based reference structures that can represent additional genetic variation from the population. In this talk, I will discuss my proposal to build on this research with a statistical algorithm for genome inference on graph-based references. This algorithm unifies several techniques from the genome inference literature in a principled statistical manner. To support the algorithm, I will generalize existing sequence-to-graph alignment concepts and build an accompanying mapping tool. I will also apply the genome inference algorithm to the problem of jointly phasing point variation and structural variation in the Simons Genome Diversity Project data set.