{"id":"W4408962005","doi":"10.1186/s40168-025-02080-3","title":"Modeling microbiome-trait associations with taxonomy-adaptive neural networks","year":2025,"lang":"en","type":"article","venue":"Microbiome","topic":"Gut microbiota and health","field":"Biochemistry, Genetics and Molecular Biology","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Microbiome; Biology; Metagenomics; Interpretability; Trait; Human Microbiome Project; Human microbiome; Computational biology; Identification (biology); Taxonomic rank; Taxon; Microbial ecology; Ecology; Evolutionary biology; Artificial intelligence; Bioinformatics; Computer science; Genetics; Gene","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.000155485,0.0002682014,0.0002601043,0.0001544046,0.0002705675,0.00006446068,0.000269299,0.0002659239,0.00001760551],"category_scores_gemma":[0.000008411873,0.0002525775,0.0001213606,0.0003843477,0.00009454579,0.000007772085,0.0001375227,0.0002053226,0.00001078279],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00009096383,"about_ca_system_score_gemma":0.0001869312,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001576272,"about_ca_topic_score_gemma":0.0003407848,"domain_scores_codex":[0.9985254,0.00006423642,0.0003064574,0.0005219464,0.00004577996,0.0005361618],"domain_scores_gemma":[0.9993004,0.000009824487,0.00008929368,0.0003628493,0.0001554447,0.00008223273],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0001030347,0.0001054314,0.0006406349,0.00002168225,0.0001786018,0.000003056284,0.00003257902,0.009857683,0.980243,0.0001179405,0.007838969,0.0008573578],"study_design_scores_gemma":[0.01926516,0.003380594,0.01863521,0.0007882391,0.001099189,0.0004655568,0.001499475,0.2078134,0.2941175,0.000215806,0.4460516,0.006668323],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8320573,0.001673564,0.1619258,0.000888856,0.0004789774,0.0009347259,0.0004717251,0.00008154193,0.001487574],"genre_scores_gemma":[0.9923896,0.00003615526,0.003229619,0.001439402,0.0001486772,0.00004736239,0.0007884831,0.00003290076,0.001887867],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6861256,"threshold_uncertainty_score":0.9999927,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01441524040319964,"score_gpt":0.240499791690801,"score_spread":0.2260845512876014,"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."}}