Decision benchmarking of cryo-EM re-refinement

Abstract

Map fit is widely used in cryo-EM re-refinement. We benchmarked selection against retaining the deposited model (no operation), using packed-side-chain agreement with higher-resolution references. In a prespecified 3.0-3.6 [A] cohort (167 entries, 131 sequence groups), highest-correlation selection increased {chi}1 disagreement by 2.4{degrees} (95% interval 2.1-2.7{degrees}), with positive loss in 91% of groups. Group-weighted fractions with >20{degrees} increases and decreases were 7.2% and 3.7%; the fraction exceeding 1 [A] local side-chain RMSD rose by 2.4 percentage points. Selection by a half map withheld from re-refinement increased disagreement (2.29{degrees}; 1.70-2.97{degrees}; 99 entries). In 53 prespecified within-study low/high-resolution pairs, loss was 3.43{degrees} (2.85-4.02{degrees}), positive in 50 pairs; half-map loss was 2.35{degrees} (1.60-3.21{degrees}; 44 pairs). Primum provides fixed cohorts, residue sets and scoring tools to benchmark refiners and selection rules against no operation.