(1) Sackler Inst. of Molecular Medicine Faculty of Medicine and
(2) Computer Science Department School of Mathematical Sciences,
Tel Aviv University, Tel Aviv 69978, Israel
(3) Laboratory of
Mathematical Biology, SAIC NCI-FCRF bldg 469, Rm 151 Frederick, MD 21702
Here we study the (in)dependence in the replacements of spatially neighboring amino acid pairs. We investigate whether the pattern of such coupled-replacements is substantially different than that of single, point interchanges. Our (Geometric Hashing) algorithm which detects similar structural motifs in proteins, affords such a study. The spatially conserved regions are found by a comparison of the coordinates of the C-alpha atoms from every pair of our structurally non-redundant dataset of proteins. This dataset has been derived from the PDB. The connectivity between the C-alpha atoms is entirely disregarded, making this technique based purely on spatial comparisons.
Here we consider only matched, spatially neighboring amino acids that are separated along the sequence by at least three residues. From these non-local neighbors in our motifs, we create a neighboring pair interchanges matrix. The size of this interchanges matrix is 400*400 (20^2*20^2). Analysis of the matrix demonstrates that in most cases the interchanges are un-coupled. That is, most of the amino acids interchange independently of their spatial neighbors. There are, however, a few exceptions to this general trend.
We have further clustered the residues in the pairs into categories. Our results indicate that in general, volume and hydrophobic/hydrophilic considerations are highly important in the coupled interchanges in the clusters.
Our study provides some guidelines in devising experiments involving point mutations, that will preserve or disrupt the stability of the protein structure. The former will be favorable interchanges in our matrix, whereas the latter will be unfavorable ones.