{"id":"W2100725988","doi":"10.1099/mic.0.2008/021360-0","title":"An analysis of initiation codon utilization in the Domain Bacteria – concerns about the quality of bacterial genome annotation","year":2008,"lang":"en","type":"article","venue":"Microbiology","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":64,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph; Public Health Agency of Canada","funders":"","keywords":"Genome; Start codon; Bacterial genome size; Codon usage bias; Biology; Bacteria; Genetics; Annotation; Computational biology; Gene; Base sequence","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":[],"consensus_categories":[],"category_scores_codex":[0.0004377067,0.00008172105,0.0002042154,0.00005499623,0.00005796714,0.000003588752,0.0001802108,0.00009569216,0.00001994081],"category_scores_gemma":[0.00003013023,0.00005520247,0.00007384642,0.0001795373,0.0002761048,0.000001168065,0.00003194606,0.00003141441,3.710886e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000007750025,"about_ca_system_score_gemma":0.00003386365,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001604503,"about_ca_topic_score_gemma":0.0009405656,"domain_scores_codex":[0.9988946,0.0004527013,0.000352854,0.0001668231,0.00003002599,0.0001029975],"domain_scores_gemma":[0.999356,0.00003884462,0.0002401958,0.0002583738,0.00009790102,0.000008694163],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"observational","study_design_scores_codex":[0.00009888501,0.00004756821,0.0298587,0.000004700679,0.0001209108,1.444167e-7,0.001734421,0.0002263451,0.9676715,0.0001605053,0.00001154821,0.00006473456],"study_design_scores_gemma":[0.0003839352,0.0002038022,0.9129728,0.000001187213,0.00006245713,0.000003391215,0.0006442551,0.00001874729,0.08381713,0.00003020526,0.001785223,0.00007685129],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9986475,0.0002957487,0.0004293073,0.00005326703,0.00006890186,0.0001561251,0.000298955,8.176977e-7,0.00004934757],"genre_scores_gemma":[0.9977428,0.0002663686,0.0001290968,0.0001040359,0.00006176763,0.00001021925,0.001677767,0.00000439391,0.000003552678],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8838544,"threshold_uncertainty_score":0.2251092,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04859194396818273,"score_gpt":0.3108261953755417,"score_spread":0.262234251407359,"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."}}