CISE News & Events
CISE PHD Student receives CSB Best Paper Award
September 4, 2008
UF CISE PhD student Ferhat Ay, his supervisor Assistant Professor Tamer Kahveci and their collaborator Assistant Professor Valerie de-Crecy Lagard (UF - Dept of Microbiology and Cell Science) received the Best Paper Award for their research paper titled "Consistent Alignment of Metabolic Pathways without Abstraction" at the International Conference of Computational Systems Biology (CSB) 2008. The CSB Conference was held in Stanford University (San Francisco), CA on August 26th - 29th, 2008. CSB is one of the most prestigious conferences in the field of Bioinformatics.
In their paper, they developed a fast and accurate algorithm for the metabolic pathway alignment problem. Metabolic pathways show complex network of reactions that transform chemical compounds. The reactions are catalyzed by biomolecules, called enzymes. Accurately aligning pathways is critical for many applications. For example, it can provide clues in drug target identification by making a comparative analysis of different organisms or improve the accuracy in phylogeny analysis.
The fundamental benefit of the algorithm in this paper lies in the innovative graph model which is able to capture all types of different interactions in between different types of nodes such as reactions, enzymes and compounds. The core part of the algorithm makes use of this model to combine the pairwise similarities of entities with the topological similarity of the query pathways. Their solution avoids the abstraction of the pathway, which is commonly used in the literature to simplify the problem at the expense of the loss in the information content. Their solution also enforces consistency in the alignment which accurately captures the biological context of the alignment process. One of the most striking aspects of their method is that it is able to identify the alternative entities and even alternative paths in milliseconds which may take weeks for biologists if done in the lab. These invaluable findings are very useful for the applications such as drug design and metabolic reconstruction. Ferhat Ay says "The performance of our algorithm is quite impressive. Especially the results being very well consistent with biological experiments are promising in the sense that our method can be used for practical purposes."
For more information and source code of this algorithm please visit:
http://bioinformatics.cise.ufl.edu/palAlign.html
Picture - From Left: Peter Markenstein (Program Co-Chair of CSB 2008), Ferhat Ay, Assistant Professor Tamer Kahveci
