Improving Contact Prediction along Three Dimensions

Abstract

Correlation patterns in multiple sequence alignments of homologous proteins can be exploited to infer information on the three-dimensional structure of their members. The typical pipeline to address this task, which we in this paper refer to as the three dimensions of contact prediction, is to (i) filter and align the raw sequence data representing the evolutionarily related proteins; (ii) choose a predictive model to describe a sequence alignment; (iii) infer the model parameters and interpret them in terms of structural properties, such as an accurate contact map. We show here that all three dimensions are important for overall prediction success. In particular, we show that it is possible to improve significantly along the second dimension by going beyond the pair-wise Potts models from statistical physics, which have hitherto been the focus of the field. These (simple) extensions are motivated by multiple sequence alignments often containing long stretches of gaps which, as a data feature, would be rather untypical for independent samples drawn from a Potts model. Using a large test set of proteins we show that the combined improvements along the three dimensions are as large as any reported to date.

Synthetic protein alignments by CCMgen quantify noise in residue-residue contact prediction

“…The gradient of the log likelihood with respect to the couplings w ij (a, b) can be written as: (11) where N ij represents the number of sequences that do not contain a gap at positions i and j, q(x i = a, x j = b) represents the empirically observed pairwise amino acid frequencies that are normalized over a, b ∈ {1, ..., 20} and 550 p(x i = a, x j = b|v, w) corresponds to the model probabilities of the MRF for observing an amino acid pair (a, b) at positions i and j. The empirical amino acid counts, given by N ij q(x i = a, x j = b), are constant and need to be computed only once from the alignment.…”

“…for residue contact prediction simply takes the L 2 norm of the 20 × 20-dimensional vector w ij with components w ij (a, b) [3,10,11,31,50],…”

From Sequence to Function: Coevolving Amino Acids Encode Structural and Functional Domains

“…Specifically, we first survey the coevolution-based profiling of functional domains for a previously-annotated set of about 800 multiple sequence alignments [43]. From this extensive survey we find that domains inferred from the sole sequence-based, coevolutionary analysis are compact in space and well consistent with the dynamical, or quasi-rigid domains inferred from the analysis of small and large-scale structural fluctuations.…”

“…Correlated substitutions can help identify those sites that host co-evolving mutations and these, in turn, are an indicator of spatial proximity [38][39][40][41][42][43][44][45][46].…”

Correlated mutations distinguish misfolded and properly folded proteins

“…Therefore, computational methods that allow for protein structure reconstruction from sequence only are greatly desired. One of these is the recently developed direct coupling analysis (DCA) method [1,2] which achieves the best results in residue-residue contact prediction from multiple sequence alignments only. Predicted contacts are used as restraints in the reconstruction of the three-dimensional structure of a protein.…”