Use of Extended Phylogenetic Profiles with E-Values and Support Vector Machines for Protein Family Classification

Protein family classification is an important means to assign functions to proteins, and use of phylogenetic profiles, which encode evolutionary history of proteins along with putative homologs, has proved to facilitate protein family classification. We proposed a new approach to compare phylogenetic profiles by incorporating the phylogenetic tree, from which the profiles are derived. Specifically, the profile is extended with new bits corresponding to the internal nodes of the tree, which encode the correlations among the bits in the original profiles. Such extension allows for direct use of E-Values, instead of imposing an ad hoc cut-off to derive binary profiles, which are commonly used in previous methods. A scoring scheme is adopted for measuring the similarity among these extended profiles, and the scores thus obtained are then provided to a classifier -a support vector machine using a polynomial kernel function -for classification. The method has been tested on the proteome of Saccharomyces cerevisiae, the budding yeast and outperformed a similar method that uses phylogenetic tree information as a tree kernel.