| Sarosh N. Fatakia | at 11:00 |
Edit |
Delete |
Affiliation: NIDDK
Computing coevolving residues in G protein-coupled receptors
G protein-coupled receptors (GPCRs) are a diverse superfamily of membrane proteins within the human genome involved in a variety of essential intercellular physiological functions and are a vital target for pharmaceutical intervention. We used mutual information (MI) across pairs of AA positions from multiple sequence alignments (MSAs) of the seven transmembrane domains (7TMs) to determine a cohort of correlated positions which may have importance for pharmaceutical intervention. We computed a MI graph in information space to represent every possible pair wise association of AA positions in the MSA. The vertices of the MI graph represented positions from the MSA and edges between them were drawn if the MI exceeded a threshold of statistical significance. Vertices with a significantly high degree lined a well defined inter-TM ligand binding cavity for class A and class C receptors but not class B for which no known ligand has been found. For a subset of class A receptors we included a portion of the second extracellular loop (EL2) and found that both positions adjacent to the conserved Cystine residue (at EL2.50) have significantly high degree. We find that the majority of these evolutionary correlated positions are under purifying selection across all subsets (obtained from the superfamily). Our algorithm may be useful for localizing topologically conserved regions involved in evolutionarily conserved function(s) in diverse protein families.
|
