Assessment of protein-protein interfaces in cryo-EM derived assemblies

Abstract

Structures of macromolecular assemblies derived from cryo-EM maps often contain errors that become more abundant with decreasing resolution. Despite efforts in the cryo-EM community to develop metrics for the map and atomistic model validation, thus far, no specific scoring metrics have been applied systematically to assess the interface between the assembly subunits. Here, we have assessed protein-protein interfaces in macromolecular assemblies derived by cryo-EM. To this end, we developed PI-score, a density-independent machine learning-based metric, trained using protein-protein interfaces’ features in high-resolution crystal structures. Using PI-score, we were able to identify errors at interfaces in the PDB-deposited cryo-EM structures (including SARS-CoV-2 complexes) and in the models submitted for cryo-EM targets in CASP13 and the EM model challenge. Some of the identified errors, especially at medium-to-low resolution structures, were not captured by density-based assessment scores. Our method can therefore provide a powerful complementary assessment tool for the increasing number of complexes solved by cryo-EM.

Interface refinement of low- to medium-resolution Cryo-EM complexes using HADDOCK2.4

“…Calculated probe scores were normalized according to the surface area (A ˚2). The recently introduced PI score was used as a second method to assess the interface quality (Malhotra et al, 2021). This method uses a support vector machine model trained on 12 interface features calculated from the atomic complex models to score each interface.…”

“…The first metric we used to assess the performance of the tested refinement protocols was the reduction in the number of b Normalized probe score representing the quality of the interface (the larger, more positive, the better), calculated using Molprobity (Word et al, 1999). c Protein interface score was calculated using the PI-score algorithm (Malhotra et al, 2021). The reported values are the average PI-score from the individual interfaces per complex.…”

“…Another metric we used to validate the quality of the PPI of the eight proposed protocols is the recently introduced PI score (Malhotra et al, 2021). Here, 12 interface features are calculated ranging from number of interface residues to shape complementary.…”