{"id":"W4415647227","doi":"10.26434/chemrxiv-2025-fz6x8","title":"The MassBank contributions of the mFam Consortium","year":2025,"lang":"en","type":"preprint","venue":"ChemRxiv","topic":"Metabolomics and Mass Spectrometry Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Identification (biology); Metadata; Annotation; Metabolomics; Metabolite; Resource (disambiguation)","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":[0.008922926,0.001966786,0.00158834,0.01070054,0.002604608,0.008101712,0.00435572,0.003277548,0.09701332],"category_scores_gemma":[0.02605993,0.000747136,0.001068275,0.01188889,0.001054948,0.009960709,0.008826843,0.002915933,0.09543711],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002365631,"about_ca_system_score_gemma":0.006217546,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003631424,"about_ca_topic_score_gemma":0.002490149,"domain_scores_codex":[0.9942115,0.0009356382,0.0007619899,0.001131561,0.002317171,0.0006421061],"domain_scores_gemma":[0.984648,0.00180599,0.001126717,0.004472504,0.005653345,0.002293419],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0004907906,0.00009697783,0.00191643,0.0005604058,0.00008021532,0.0003110244,0.0001714352,0.0003336175,0.002732392,0.02505255,0.7879688,0.1802854],"study_design_scores_gemma":[0.00003202555,0.00001381418,0.001131752,0.0001626218,0.00002925391,0.0002444285,0.00004742106,0.0007769414,0.001067101,0.006858668,0.9896095,0.00002637968],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"other","genre_gemma":"dataset","genre_scores_codex":[0.01622377,0.02399138,0.1031532,0.0921693,0.01771757,0.001190235,0.3274946,0.0659894,0.3520705],"genre_scores_gemma":[0.04083187,0.01633723,0.1087361,0.01613964,0.006692349,0.00180749,0.6495111,0.01344128,0.1465029],"genre_candidate":"dataset","genre_consensus":null,"teacher_disagreement_score":0.09701332,"threshold_uncertainty_score":0.3245419,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.008766325145400714,"score_gpt":0.26315525433525,"score_spread":0.2543889291898493,"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."}}