Protein side-chain packing problem: is there still room for improvement?

Abstract

The protein side-chain packing problem (PSCPP) is an important subproblem of both protein structure prediction and protein design. During the past two decades, a large number of methods have been proposed to tackle this problem. These methods consist of three main components: a rotamer library, a scoring function and a search strategy. The average overall accuracy level obtained by these methods is approximately 87%. Whether a better accuracy level could be achieved remains to be answered. To address this question, we calculated the maximum accuracy level attainable using a simple rotamer library, independently of the energy function or the search method. Using 2883 different structures from the Protein Data Bank, we compared this accuracy level with the accuracy level of five state-of-the-art methods. These comparisons indicated that, for buried residues in the protein, we are already close to the best possible accuracy results. In addition, for exposed residues, we found that a significant gap exists between the possible improvement and the maximum accuracy level achievable with current methods. After determining that an improvement is possible, the next step is to understand what limitations are preventing us from obtaining such an improvement. Previous works on protein structure prediction and protein design have shown that scoring function inaccuracies may represent the main obstacle to achieving better results for these problems. To show that the same is true for the PSCPP, we evaluated the quality of two scoring functions used by some state-of-the-art algorithms. Our results indicate that neither of these scoring functions can guide the search method correctly, thereby reinforcing the idea that efforts to solve the PSCPP must also focus on developing better scoring functions.

Antibody side chain conformations are position‐dependent

“…During side chain prediction, the torsion angles of the side chains, the χ angle(s), must be determined using only the existing backbone structure . This is often done by describing the series of χ angles as discrete forms, known as rotamers .…”

OPUS-Rota4: a gradient-based protein side-chain modeling framework assisted by deep learning-based predictors

“…As shown in Table 6 , OPUS-Fold2 can guide the side chains to their proper places with the correct side-chain contact constraints, showing the effectiveness of its side-chain modeling ability. In this case, we can improve the protein side-chain modeling accuracy by improving the accuracy of side-chain contact map prediction other than developing better scoring functions [ 16 ].…”

“…Since the backbone modeling driven by backbone contact map works well, we can develop the side-chain contact map for side-chain modeling accordingly. In this case, we convert the protein side-chain modeling problem from developing better scoring functions [ 16 ] to improving the accuracy of side-chain contact map prediction.…”

Scoring of Side-Chain Packings: An Analysis of Weight Factors and Molecular Dynamics Structures

“…A group of sampled conformations from this simulation is used as a reference, instead of the crystal structures, to assess the performance of the scoring functions of two state-of-the-art methods for the PSCPP, using a methodology previously proposed. 31 ■ MATERIALS AND METHODS Quality Measures. The methods for PSCPP can be evaluated in silico, contrasting their predictions against the experimental structures stored in the Protein Data Bank (PDB).…”

“…8−11 Furthermore, it is sufficient for achieving accuracies near the ideal values. 31 The rotamers in the library are sorted from highest to lowest probability, and the library is truncated to those rotamers with cumulative probability of 0.98, as considered in some previous works. 4,10 Data Sets.…”