Classification of Transmembrane Protein Families Based on Topological Similarity

摘要

Transemembrane (TM) proteins occupies about one fourth of proteome in each species, playing essential role in life. Nevertheless, more than one half of them are still left function-unknown or not well annotated. It is because the sequence-similarity based search is less effective in the functional assignment for TM proteins comparing to soluble proteins. This is attributed mostly to their structural making that their polypeptide chain comprises of several hydrophobic regions (TM segments) with much less conservative ammo acid sequence than soluble region. Our aim is to classify TM protein functions more comprehensively and more in detail based on topology similarity in addition to usual sequence similarity. In this study, we focus on 4-tms TM proteins of various functions as an example to show the use of topology similarity improves significantly not only the functional classification of TM proteins but also the functional annotation for TM protein sequences with unknown function.