HONG KONG BAPTIST UNIVERSITY
FACULTY OF SCIENCE
Department of Computer Science Colloquium
Surprising Results on Phylogenetic Tree Building Methods Based on Molecular Sequences
Prof. Gaston Gonnet
Head of Computational Biology and Bioinformatics
Swiss Federal Institute of Technology
Date: February 7, 2013 (Thursday)
Time: 10:30 - 11:30 am
Venue: SCT909, Cha Chi Ming Science Tower, Ho Sin Hang Campus
The problem of building phylogenetic trees is of central importance in bioinformatics. It is important as an end in several applications and it is also a tool on which many other processes are based on. Hence the importance of building good trees. On the other hand, bioinformatics has to provide answers to problems for which we do not have an oracle, that is, we do not really know the answer. Verifying that the methods are accurate and comparing methods is difficult and suffers from several flaws, most of them related to the classical problems of model selection (if the model is not accurate the comparisons may not be meaningful).
We develop measures of accuracy which have biological roots and are model independent. This allows us to determine the quality of the tree building methods with more precision. The results are quite surprising for the community.
Professor Gonnet started his academic career in the analysis of algorithms. Later, in 1980, together with Keith Geddes, he formed the Symbolic Computation Group, a group devoted to research in Symbolic Computation or Computer Algebra, and to the development of the Maple Algebra System. Maple has found its way into the practical world, aiding engineers to do their computations, assisting scientists in their research and helping students to learn mathematics.
In 1983-84, the University of Waterloo and Oxford University Press became partners for the computerization of the Oxford English Dictionary. At that time, Professor Gonnet and colleagues founded the "Centre for the New Oxford English Dictionary''. The Centre has attracted a lot of activity around the work on the dictionary and also in connection with the research being done with large text databases. Some of these activities reached their climax with the publication of the second edition of the dictionary, a work which would have not been possible without intelligent text processing. The main contributions of this project have been in the areas of fast text searching, text structuring and text transformations which then were reused in bioinformatics.
In 1989, Professor Gonnet accepted a position with E.T.H. Zurich, where he is working in Bioinformatics. Professor Gonnet and Prof Steven Benner founded the ETH Computational Biochemistry Research Group. The CBRG was responsible for the first self-matching of an entire protein database. The CBRG specializes in sequence analysis, models of evolution and phylogeny construction. The work in bioinformatics has been extended to various aspects including the creation and curation in 2005 of the OMA database, the largest database of orthologous relations. The CBRG is a member of the Swiss Institute of Bioinformatics.
In 1989, Professor Gonnet was awarded the Information Technology Association of Canada annual award for his contributions to computer algebra and text searching. In 2012 he was awarded a Dr. Honoris Causa by the Universidad de la Republica in Uruguay for the various contributions to computer science.
Professor Gonnet has started several companies related to his research, most notably Open Text and Waterloo Maple, both companies located in Canada.
********* ALL INTERESTED ARE WELCOME ***********
(For enquiry, please contact Computer Science Department at 3411 2385)
Department of Computer Science, Hong Kong Baptist University