- Candidate for Master of Science in Biology (Computational Biology), McMaster University, 2007 - present
- Bachelor of Computer Science in Honours Computer Science - Bioinformatics Option, Co-operative Program, University of Waterloo, 2002 - 2007
I am currently interested gene fusion and gene fission events that occur in bacterial genomes.[Back to top]
- Biology 721 - Topics in Molecular Evolution
- Biology 724 - Molecular Ecology
- Biology 2F03 - Fundamental & Applied Ecology
- Biology 3CC3 - Microbial Genomes & Systems Biology
In undergraduate courses I took at the University of Waterloo, WebPHYLIP was used in labs and assignments for phylogenetic inference. However, it was cumbersome and web servers hosting WebPHYLIP were becoming increasingly scarce. Thus, as a forth year bioinformatics course project, I worked on a two person team, with Eric Chiang, with the goal of developing a proof of concept website for phylogenetic inference that is easy to use and targetted at teaching at the undergraduate level.
Web server features:
- Written entirely in Python and is self-contained, including server-side web scripting and inference algorithms (i.e. not a web interface wrapped around the PHYLIP package).
- Automatic parsing of multiple sequence alignment output from ClustalW and TCoffee (no pre-processing required).
Three types of inference methods implemented:
- Unweighted Pair Group Method with Arithmetic mean (UPGMA)
- Neighbour Joining (NJ)
- Maximum parsimony (based on the Sankoff algorithm)
- Outputs distance matrix, text tree in Newick format, and graphical tree in PNG format for UPGMA and NJ.
- Outputs Newick tree only for maximum parsimony.
As it was a proof of concept only, particularly due to time constraints, please keep the following in mind:
- Correctness of the algorithms cannot be guaranteed (there was no time for extensive testing).
- Only DNA sequences are recognized.
- Only the Jukes-Cantor model of nucleotide substitution is available.
- The maximum parsimony algorithm was written in a brute-force approach with no optimizations whatsoever (exponential runtime due to double recursions for iterating over all possible trees).
- The Golding Lab
- Computational Biology at McMaster University
- Deptartment of Biology, McMaster University
- McMaster University