|M.Sc Student||Mano Adi|
|Subject||Building Phylogenetic Trees Based on Biochemical Pathways|
|Department||Department of Computer Science||Supervisor||Professor Ron Pinter|
|Full Thesis text|
Evolutionary information can be derived from many kinds of data. One notable possibility is the comparison among known pathways of several organisms. Using such data judiciously enhances the current arsenal of phylogeny measures (such as genomic data, Ribosomal RNA 16S sequences, and more), elucidates the evolutionary history of specific pathways and of the metabolic systems of organisms as a whole, and - in a complementary manner - provides the means to hone and improve pathway alignment scoring methods and the quality of data in existing repositories. In this work I present a new computational pipeline that builds phylogenies based on merging the comparison of multiple metabolic pathways and the results of applying it to data from the MetaCyc repository. This presents a significant computational challenge, which was solved by parallelizing the computation using a grid facility. Some interesting conclusions can be made about the parameters used within this pipeline as well as for the alignment algorithm I used to compare the individual metabolic pathways. Furthermore, my results reveal several interesting biological observations about certain metabolic similarities in Archaea. Finally, they reflect on the quality of the information that is currently retrieved in on-line sources, which I found to be lacking in completeness as well as in accuracy.