{"id":"W4405778687","doi":"10.1101/2024.12.15.628585","title":"EnzymeCAGE: A Geometric Foundation Model for Enzyme Retrieval with Evolutionary Insights","year":2024,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Microbial Metabolic Engineering and Bioproduction","field":"Biochemistry, Genetics and Molecular Biology","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Mila - Quebec Artificial Intelligence Institute","funders":"","keywords":"Foundation (evidence); Computer science; Artificial intelligence; Geography; Archaeology","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004641181,0.000875819,0.0007700866,0.0007346075,0.000357583,0.0009852622,0.001647764,0.001293428,0.003779779],"category_scores_gemma":[0.00174863,0.0004645116,0.001347374,0.0005602601,0.0005485365,0.001123197,0.0008743657,0.001123213,0.001012391],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008907213,"about_ca_system_score_gemma":0.001061844,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006913326,"about_ca_topic_score_gemma":0.0079879,"domain_scores_codex":[0.9998443,0.00004103954,0.000008514534,0.00004723488,0.00003489637,0.00002405478],"domain_scores_gemma":[0.9995846,0.000194427,0.00003753328,0.00007517119,0.00007017351,0.00003812649],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0000896848,0.00003031416,0.001086981,0.00004464652,0.00003645277,0.00006494648,0.00002168659,0.9667453,0.001712066,0.00959434,0.00187287,0.01870078],"study_design_scores_gemma":[0.000005778637,0.00000609736,0.00002977978,0.000001919474,0.000001866048,0.000007955035,0.000002078794,0.997195,0.0002255781,0.00212311,0.0003989064,0.000001924205],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1113972,0.0006467869,0.8748541,0.000539123,0.00008974371,0.00008253807,0.002046043,0.004431433,0.005912968],"genre_scores_gemma":[0.7022607,0.000467965,0.2845432,0.0003297249,0.00008082792,0.0002928868,0.006756828,0.0008589253,0.004408851],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006913326,"threshold_uncertainty_score":0.0137462,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.009832383824401011,"score_gpt":0.2099745982871137,"score_spread":0.2001422144627126,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}