A Pareto-optimal approach for protein structure evaluation using Amber and Rosetta energy functions.

An accurate energy function is an essential component of biomolecular structural modeling and design. The comparison of differently derived energy functions enables analysis of the strengths and weaknesses of each energy function, and provides independent benchmarks for evaluating improvements within a given energy function. We compared the molecular mechanics Amber empirical energy function to two versions of the Rosetta energy function (talaris2014 and REF2015) in decoy discrimination and loop modeling tests. Both Rosetta's talaris2014 and Amber's ff14SBonlySC energy functions performed well in scoring the native state as the lowest energy conformation in many cases. In 24/150 cases with Rosetta, and in 2/150 cases using Amber, a false minimum is found that is absent in the alternative landscape. In 21/150 cases, both energy function-generated landscapes featured false minima. The newest version of the Rosetta energy function, REF2015, which has more physically-derived terms than talaris2014, performs significantly better, highlighting the improvements made to the Rosetta scoring approach. To take advantage of the semi-orthogonal nature of these energy functions, we developed a Pareto optimization approach that combines Amber and Rosetta energy landscapes to predict the most near-native model for a given protein. This algorithm improves upon predictions from either energy function in isolation, and should aid in model selection for structure prediction and loop modeling tasks.