*Q-score as a reliability measure for protein, nucleic acid and
small-molecule atomic coordinate models derived from 3DEM maps*
Grigore Pintilie, Chenghua Shao, Zhe Wang, Brian P Hudson, Justin W
Flatt, Michael F Schmid, Kyle L Morris, Stephen K Burley, Wah Chiu
(2025) /Acta Cryst/ D81: 410-422 doi: 10.1107/S2059798325005923
<https://doi.org/10.1107/S2059798325005923> 

A new publication describes two percentile-based metrics: Q_relative_all and Q_relative_resolution. The publication presents a statistical analysis of Q-score and these new percentile based validation metrics. *Q_relative_all* is a single value percentile, representing the overall quality of the model-map fitness. It is derived by comparing a model-map average Q-score to all model-map average Q-scores calculated for 3DEM structures in the PDB/EMDB. Higher Q_relative_all percentile values, infer higher quality in the model-map fitness. *Q_relative_resolution* is a single value percentile, derived by comparing the average Q-score of a model-map with the Q-scores of other model-map entries of similar resolution. This metric was developed from the principle that average Q-score is correlated with map resolution. Q_relative_resolution percentile values close to 50% therefore represent a score that is commonly observed for entries in the PDB/EMDB of similar resolution, inferring model-map fitness is typical for an entry of a particular resolution. Notably low or high Q_relative_resolution percentiles may be helpful to guide further review of model-maps. The EMDB has made these metrics available since Spring 2024 via emdatabank.org <https://emdatabank.org> and since the publication of the paper, the wwPDB is in the process of incorporating these metrics into wwPDB validation reports by providing the Q-score percentile slider for the assessment of model-map fitness of 3DEM entries. This validation enhancement will help depositors, reviewers and the community to improve the quality of data submissions and assist users to assess 3DEM data quality in the PDB and EMDB, wwPDB core archives. *Q-score as a reliability measure for protein, nucleic acid and small-molecule atomic coordinate models derived from 3DEM maps* Grigore Pintilie, Chenghua Shao, Zhe Wang, Brian P Hudson, Justin W Flatt, Michael F Schmid, Kyle L Morris, Stephen K Burley, Wah Chiu (2025) /Acta Cryst/ D81: 410-422 doi: 10.1107/S2059798325005923 <https://doi.org/10.1107/S2059798325005923> Please read the full news here: https://www.wwpdb.org/news/news?year=2025#689f2e8a3b59581b680197a2 On behalf of the wwPDB -- IRINA PERSIKOVA, Ph.D. Deputy Biocuration Lead, RCSB Protein Data Bank Research Associate, Institute for Quantitative Biomedicine Rutgers, The State University of New Jersey 174 Frelinghuysen Road, Piscataway NJ 08854 P: 848.445.4938 | E: irina.persikova@rcsb.org <mailto:irina.persikova@rcsb.org> rcsb.org <www.rcsb.org> | iqb.rutgers.edu <http://iqb.rutgers.edu> | facebook <https://www.facebook.com/RCSBPDB> | twitter <https://twitter.com/buildmodels>