Fast Homology Search using Categorisation Profiles

Homology search is an important step in discovering evolutionary relationships in modern molecular biology. In particular, it is key to analysing data from large-scale sequencing initiatives: by establishing homology between a newly-discovered sequence and well-understood sequences in curated repositories, it is possible to infer structure and function without the need for costly, time-consuming wet laboratory work. Biologists almost always search protein collections in preference to nucleotide collections because, in general, proteins are better annotated and, therefore, more useful in discovering evolutionary relationships. It is possible to deduce the function of a protein based on its similarity to sequences that have been previously characterised by comparing the primary or linear structure of a protein. Indeed, more than half of the data derived from newly-sequenced genomes can be characterised based on its similarity to other well-understood organisms [5]. As new genomes are sequenced and annotated, function elucidation for this new data becomes increasingly possible.