BAYESIAN STATISTICAL-ANALYSIS OF PROTEIN SIDE-CHAIN ROTAMER PREFERENCES

We present a Bayesian statistical analysis of the conformations of sid e chains in proteins from the Protein Data Bank. This is an extension of the backbone-dependent rotamer library, and includes rotamer popula tions and average chi angles for a full range of phi,psi values. The B ayesian analysis used here provides a rigorous statistical method for taking account of varying amounts of data. Bayesian statistics require s the assumption of a prior distribution for parameters over their ran ge of possible values. This prior distribution can be derived from pre vious data or from pooling some of the present data. The prior distrib ution is combined with the data to form the posterior distribution, wh ich is a compromise between the prior distribution and the data. For t he chi(2), chi(3), and chi(4) rotamer prior distributions, we assume t hat the probability of each rotamer type is dependent only on the prev ious chi rotamer in the chain. For the backbone-dependence of the chi( 1) rotamers, we derive prior distributions from the product of the phi -dependent and psi-dependent probabilities. Molecular mechanics calcul ations with the CHARMM22 potential show a strong similarity with the e xperimental distributions, indicating that proteins attain their lowes t energy rotamers with respect to local backbone-side-chain interactio ns. The new library is suitable for use in homology modeling, protein folding simulations, and the refinement of X-ray and NMR structures.