{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,24]],"date-time":"2026-04-24T15:15:23Z","timestamp":1777043723087,"version":"3.51.4"},"reference-count":25,"publisher":"Oxford University Press (OUP)","issue":"4","license":[{"start":{"date-parts":[[2026,4,2]],"date-time":"2026-04-02T00:00:00Z","timestamp":1775088000000},"content-version":"vor","delay-in-days":1,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Machine Learning for Pharmaceutical Discovery and Synthesis"},{"name":"NSF Expeditions","award":["1918839"],"award-info":[{"award-number":["1918839"]}]},{"name":"Abdul Latif Jameel Clinic"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,4,7]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Motivation<\/jats:title>\n                    <jats:p>Protein structure generative models have seen a recent surge of interest, but meaningfully evaluating them computationally is an active area of research. While current metrics have driven useful progress, they do not capture how well models sample the design space represented by the training data. We argue for a protein Frechet Inception Distance (FID) metric to supplement current evaluations with a measure of distributional similarity in a semantically meaningful latent space.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Our FID behaves desirably under protein structure perturbations and correctly recapitulates similarities between protein samples: it correlates with optimal transport distances and recovers FoldSeek clusters and the CATH hierarchy. Evaluating current protein structure generative models with FID shows that they fall short of modeling the distribution of PDB proteins.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Availability<\/jats:title>\n                    <jats:p>Code is available at: https:\/\/github.com\/ffaltings\/protfid.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1093\/bioinformatics\/btag156","type":"journal-article","created":{"date-parts":[[2026,4,1]],"date-time":"2026-04-01T11:45:29Z","timestamp":1775043929000},"source":"Crossref","is-referenced-by-count":0,"title":["Protein FID: improved evaluation of protein structure generative models"],"prefix":"10.1093","volume":"42","author":[{"ORCID":"https:\/\/orcid.org\/0009-0002-8424-7467","authenticated-orcid":false,"given":"Felix","family":"Faltings","sequence":"first","affiliation":[{"name":"CSAIL, MIT , Cambridge, MA 02139,","place":["United States"]}]},{"given":"Hannes","family":"Stark","sequence":"additional","affiliation":[{"name":"CSAIL, MIT , Cambridge, MA 02139,","place":["United States"]}]},{"given":"Tommi","family":"Jaakkola","sequence":"additional","affiliation":[{"name":"CSAIL, MIT , Cambridge, MA 02139,","place":["United States"]}]},{"given":"Regina","family":"Barzilay","sequence":"additional","affiliation":[{"name":"CSAIL, MIT , Cambridge, MA 02139,","place":["United States"]}]}],"member":"286","published-online":{"date-parts":[[2026,4,2]]},"reference":[{"key":"2026042409462503200_btag156-B1","doi-asserted-by":"crossref","first-page":"493","DOI":"10.1038\/s41586-024-07487-w","article-title":"Accurate structure prediction of biomolecular interactions with alphafold 3","volume":"630","author":"Abramson","year":"2024","journal-title":"Nature"},{"key":"2026042409462503200_btag156-B2","first-page":"1514","volume-title":"Nat Methods","author":"Ahdritz"},{"key":"2026042409462503200_btag156-B3","author":"Bose","year":"2023"},{"key":"2026042409462503200_btag156-B5","author":"Campbell","year":"2024"},{"key":"2026042409462503200_btag156-B6","doi-asserted-by":"crossref","first-page":"49","DOI":"10.1126\/science.add2187","article-title":"Robust deep learning\u2013based protein sequence design using proteinmpnn","volume":"378","author":"Dauparas","year":"2022","journal-title":"Sci"},{"key":"2026042409462503200_btag156-B7","author":"Geffner"},{"key":"2026042409462503200_btag156-B8","first-page":"2252","volume-title":"Nat Struct Mol Biol","author":"Glasscock","year":"2023"},{"key":"2026042409462503200_btag156-B9","first-page":"850","volume-title":"Sci","author":"Hayes"},{"key":"2026042409462503200_btag156-B10","article-title":"GANs trained by a two time-scale update rule converge to a local nash equilibrium","volume":"30","author":"Heusel","year":"2017","journal-title":"Adv Neural Inf Process Syst"},{"key":"2026042409462503200_btag156-B11","doi-asserted-by":"crossref","first-page":"e0199585","DOI":"10.1371\/journal.pone.0199585","article-title":"Contact prediction is hardest for the most informative contacts, but improves with the incorporation of contact potentials","volume":"13","author":"Holland","year":"2018","journal-title":"PLoS One"},{"key":"2026042409462503200_btag156-B12","doi-asserted-by":"crossref","first-page":"1070","DOI":"10.1038\/s41586-023-06728-8","article-title":"Illuminating protein space with a programmable generative model","volume":"623","author":"Ingraham","year":"2023","journal-title":"Nature"},{"key":"2026042409462503200_btag156-B13","author":"Kynk\u00e4\u00e4nniemi","year":"2022"},{"key":"2026042409462503200_btag156-B14","author":"Lin","year":"2023"},{"key":"2026042409462503200_btag156-B15","author":"Lin","year":"2024"},{"key":"2026042409462503200_btag156-B16","doi-asserted-by":"crossref","first-page":"1123","DOI":"10.1126\/science.ade2574","article-title":"Evolutionary-scale prediction of atomic-level protein structure with a language model","volume":"379","author":"Lin","year":"2023","journal-title":"Sci"},{"key":"2026042409462503200_btag156-B17","author":"Lu"},{"key":"2026042409462503200_btag156-B18","doi-asserted-by":"crossref","first-page":"E7438","DOI":"10.1073\/pnas.1607178113","article-title":"Tertiary alphabet for the observable protein structural universe","volume":"113","author":"Mackenzie","year":"2016","journal-title":"Proc Natl Acad Sci U S A"},{"key":"2026042409462503200_btag156-B19","doi-asserted-by":"crossref","first-page":"1093","DOI":"10.1016\/S0969-2126(97)00260-8","article-title":"Cath\u2013a hierarchic classification of protein domain structures","volume":"5","author":"Orengo","year":"1997","journal-title":"Struct"},{"key":"2026042409462503200_btag156-B20","doi-asserted-by":"crossref","first-page":"243","DOI":"10.1038\/s41587-023-01773-0","article-title":"Fast and accurate protein structure search with foldseek","volume":"42","author":"Van Kempen","year":"2024","journal-title":"Nat Biotechnol"},{"key":"2026042409462503200_btag156-B21","doi-asserted-by":"crossref","first-page":"1089","DOI":"10.1038\/s41586-023-06415-8","article-title":"De novo design of protein structure and function with rfdiffusion","volume":"620","author":"Watson","year":"2023","journal-title":"Nature"},{"key":"2026042409462503200_btag156-B22","author":"Yim","year":"2023"},{"key":"2026042409462503200_btag156-B23","author":"Yim"},{"key":"2026042409462503200_btag156-B24","doi-asserted-by":"crossref","first-page":"702","DOI":"10.1002\/prot.20264","article-title":"Scoring function for automated assessment of protein structure template quality","volume":"57","author":"Zhang","year":"2004","journal-title":"Proteins Struct Funct Bioinforma"},{"key":"2026042409462503200_btag156-B25","author":"Zhang"},{"key":"2026042409462503200_btag156-B26","author":"Zhou","year":"2020"}],"container-title":["Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bioinformatics\/advance-article-pdf\/doi\/10.1093\/bioinformatics\/btag156\/67726836\/btag156.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/42\/4\/btag156\/67726836\/btag156.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/42\/4\/btag156\/67726836\/btag156.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,24]],"date-time":"2026-04-24T13:46:33Z","timestamp":1777038393000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article\/doi\/10.1093\/bioinformatics\/btag156\/8572629"}},"subtitle":[],"editor":[{"given":"Lenore","family":"Cowen","sequence":"additional","affiliation":[]}],"short-title":[],"issued":{"date-parts":[[2026,4]]},"references-count":25,"journal-issue":{"issue":"4","published-print":{"date-parts":[[2026,4,7]]}},"URL":"https:\/\/doi.org\/10.1093\/bioinformatics\/btag156","relation":{},"ISSN":["1367-4803","1367-4811"],"issn-type":[{"value":"1367-4803","type":"print"},{"value":"1367-4811","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2026,4]]},"published":{"date-parts":[[2026,4]]},"article-number":"btag156"}}