{"id":"W4416947312","doi":"10.1038/s41597-025-06191-2","title":"Reconstruction of 1,979 prokaryotic metagenome-assembled genomes from 37 global cave environments","year":2025,"lang":"en","type":"article","venue":"Scientific Data","topic":"Subterranean biodiversity and taxonomy","field":"Earth and Planetary Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Trinity Western University; Western University","funders":"Natural Science Foundation of Guangdong Province; National Natural Science Foundation of China; Li Ka Shing Foundation","keywords":"Cave; Phylum; Genome; Metagenomics; Archaea; Phylogenetics; Extremophile; Gene; Biodiversity","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0004387355,0.0001106998,0.0001755388,0.00008277387,0.0002733743,0.0001450958,0.001063955,0.00005397775,0.003829019],"category_scores_gemma":[0.00003158802,0.0001021498,0.0000424643,0.0004103765,0.0002713672,0.000522388,0.0001854331,0.00006462326,0.0005510806],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001396738,"about_ca_system_score_gemma":0.00008838153,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001444154,"about_ca_topic_score_gemma":0.001474798,"domain_scores_codex":[0.9986,0.00006926671,0.0002263273,0.0006404474,0.0002461802,0.0002177705],"domain_scores_gemma":[0.9987192,0.00005361232,0.000109657,0.001028256,0.00001414017,0.00007513749],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00003484301,0.00002914832,0.8461586,0.00001376704,0.0001097708,0.000005248564,0.00005568129,0.00005608554,0.0002155096,0.00001585967,0.004953036,0.1483524],"study_design_scores_gemma":[0.0005468951,0.00002732713,0.9027177,0.00003047299,0.0001673061,0.000004780661,0.0005984465,0.003503475,0.0009352674,0.00147008,0.08978368,0.0002145364],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9590502,0.00165312,0.0005587836,0.00009908532,0.003645171,0.0003249498,0.01851763,0.00002733841,0.01612366],"genre_scores_gemma":[0.988245,0.00003824495,0.003606941,0.00004753377,0.00003197763,4.101809e-7,0.006863086,9.442494e-7,0.001165937],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1481379,"threshold_uncertainty_score":0.9970816,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05960361819547653,"score_gpt":0.2232549988934293,"score_spread":0.1636513806979528,"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."}}