Per-residue optimisation of protein structures: rapid alternative to local optimisation with constrained alpha carbons

Abstract

In recent years, the number of known protein structures has increased significantly. Predictive algorithms and experimental methods provide the global topology and fold of proteins with high accuracy. However, the local quality of the protein structures, including bond lengths, angles, and positions of individual atoms, often lacks the same level of precision due to, for example, the accuracy of the predictive algorithms or the resolution of the experiments. For this reason, protein structures are usually optimised by a force field prior to their application in further research sensitive to structural quality. Protein structure optimisation, however, is computationally challenging. In this paper, we introduce a method Per-residue optimisation of protein structures: Rapid alternative to local optimisation with constrained alpha carbons (PROPTIMUS RAPHAN), which is applicable to any force field. Rather than optimising the entire protein structure at once, PROPTIMUS RAPHAN divides the structure into overlapping residual substructures and optimises each substructure individually. This approach results in computational time that scales linearly with the size of the structure. Additionally, we present PROPTIMUS RAPHAN \(_{\text {GFN-FF}}\) , a reference implementation of our method employing a generic, almost QM-accurate force field, GFN-FF. We tested PROPTIMUS RAPHAN \(_{\text {GFN-FF}}\) on 461 AlphaFold DB structures and demonstrated that our approach achieves results comparable to the optimisation of the structure with constrained alpha carbons in significantly less time. Scientific contribution The main contribution of this work is the PROPTIMUS RAPHAN method and its reference parallelisable implementation PROPTIMUS RAPHAN \(_{\text {GFN-FF}}\) . Because the time requirement increases linearly with the size of the structure, PROPTIMUS RAPHAN \(_{\text {GFN-FF}}\) optimises on average 5,000 atoms per hour on the CPU used for testing. Therefore, prior to any research sensitive to protein structure quality, our method can be employed to obtain protein structures closer to QM-accuracy.