{"id":"W3009070426","doi":"10.1099/mgen.0.000341","title":"An assessment of genome annotation coverage across the bacterial tree of life","year":2020,"lang":"en","type":"article","venue":"Microbial Genomics","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":122,"is_retracted":false,"has_abstract":true,"ca_institutions":"Wilfrid Laurier University; University of Waterloo","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Annotation; Genome; Proteome; Biology; Bacterial genome size; Computational biology; Homology (biology); Genome project; Gene Annotation; Gene; Phylogenetics; Genome size; Genetics","routes":{"ca_aff":true,"ca_fund":true,"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":[0.00683344,0.0004559424,0.0006877774,0.005375257,0.0006249102,0.001766884,0.0005348784,0.0005051846,0.000640173],"category_scores_gemma":[0.02106686,0.0002715424,0.0006355685,0.004188757,0.0005517156,0.001785771,0.001598815,0.0004859141,0.0003400973],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003852098,"about_ca_system_score_gemma":0.000561539,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001349342,"about_ca_topic_score_gemma":0.001587521,"domain_scores_codex":[0.9960352,0.001311807,0.0004404745,0.0006803023,0.001103518,0.0004287492],"domain_scores_gemma":[0.9789973,0.008415918,0.004947628,0.001646133,0.005225079,0.0007679862],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0009063825,0.00005057044,0.84574,0.000746947,0.0006446099,0.0001829748,0.00203398,0.003503656,0.09609839,0.0007073841,0.0004139881,0.04897125],"study_design_scores_gemma":[0.000005216973,0.0002607107,0.9803628,0.00007952601,0.0001602962,0.000325848,0.0008319579,0.005352823,0.008627698,0.0006199513,0.003340732,0.00003250409],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9889181,0.002365168,0.005477982,0.0001352464,0.000008424412,0.00002046575,0.001883105,0.00009907416,0.001092438],"genre_scores_gemma":[0.9841188,0.0009692526,0.008193321,0.00007631444,0.00001441913,0.00004513933,0.006231488,0.00005455258,0.0002966636],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.00683344,"threshold_uncertainty_score":0.03613907,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01321447599047545,"score_gpt":0.2681903846297509,"score_spread":0.2549759086392755,"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."}}