- This event has passed.
BC Cancer Research Seminar Series: Dr. Jens Stoye
March 9 @ 12:00 pm - 1:00 pm
BC Cancer Research Seminar Series presents a seminar:
Dr. Jens Stoye, PhD
Professor, Bioinformatics, Bielefeld University, Germany, University of California, German Cancer Research Center in Heidelberg, The Max Planck Institute for Molecular Genetics in Berlin
Date: Monday, March 9, 2020
Time: 12:00pm to 1:00pm
Location: Gordon & Leslie Diamond Family Theatre, BC Cancer Research Centre, 675 W.10th Ave.
Videolinked to: FVC – room 2035, VIC – room #3, CSI – Shuswap room, CN – Skeena room, UBC Biomedical Research Centre – room 351, Michael Smith Labs – room 303, C&W – room K4-142
Title: Detecting High Scoring Local Alignments in Colored de Bruijn Graphs
Increasing amounts of individual genomes sequenced per species motivate the usage of pan genomic approaches. Pan genomes may be represented as graphical structures, e.g. compacted colored de Bruijn graphs, which offer a low memory usage and facilitate reference-free sequence comparisons. While sequence-to-graph alignment on graphical pan genomes has been studied for some time, no local alignment search tool in the vein of BLAST has been proposed yet.
We developed a new method to find all maximum scoring local alignments between a DNA query sequence and a pan genome represented as a compacted colored de Bruijn graph. In addition to the mere comparison of query and pan genome sequences, it also allows a comparison of similarity among sequences within the pan genome. We demonstrate that the theory of pairwise sequence alignment statistics generalizes to sequence-to-graph alignments and use it to separate local alignments representing sequence homology from spurious findings. Furthermore, we introduce the notions of quorum and search color set allowing concentrating searches on any part of the pan genome without reconstructing the graph.
An implementation of our method is presented, and its performance and usability are shown. Our approach scales sub linearly in running time and memory usage with respect to the number of genomes under consideration. Thus, it is superior to classical methods that do not make use of sequence similarity within the pan genome.
This is joint work with Tizian Schulz, Faraz Hach, Sven Rahmann and Roland Wittler.
Jens Stoye received his PhD (1997) in Bioinformatics from Bielefeld University, Germany. After postdoctoral positions at the University of California at Davis (1997-1998) and the German Cancer Research Center in Heidelberg (1998-2001), he became head of the Algorithmic Bioinformatics group at the Max Planck Institute for Molecular Genetics in Berlin (2001-2002). Stoye has been a full professor for Genome Informatics back at Bielefeld University since 2002. His research interests are in algorithms for bioinformatics, genome-scale sequence analysis, and comparative genomics.
Sponsored by: BC Cancer Foundation
Hosted by: Dr. Inanc Birol