{"id":"W2137516483","doi":"10.1371/journal.pbio.1001920","title":"Genomic Encyclopedia of Bacteria and Archaea: Sequencing a Myriad of Type Strains","year":2014,"lang":"en","type":"article","venue":"PLoS Biology","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":234,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"Natural Environment Research Council; Office of Science; Agence Nationale de la Recherche; U.S. Department of Energy","keywords":"Biology; Archaea; Genetics; Genetic diversity; Genome; Evolutionary biology; Diversity (politics); Metagenomics; Function (biology); Gene; Genomics; Encyclopedia; Horizontal gene transfer; Bacteria; Computational biology; Library science; Population","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":[0.002018012,0.002611517,0.002313774,0.0106564,0.0009674496,0.003774488,0.001875275,0.001049146,0.02302446],"category_scores_gemma":[0.006212378,0.0007812722,0.001080149,0.02149394,0.0006870961,0.001682724,0.001915271,0.002103631,0.02737874],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006096608,"about_ca_system_score_gemma":0.004560173,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003704306,"about_ca_topic_score_gemma":0.004950075,"domain_scores_codex":[0.9983747,0.0002507667,0.0004534237,0.0003543568,0.0004712964,0.00009542004],"domain_scores_gemma":[0.9952859,0.001330323,0.0009058529,0.0009766446,0.0009690988,0.000532091],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"observational","study_design_scores_codex":[0.001484497,0.0002457324,0.01181918,0.02399243,0.0007090906,0.001976939,0.001160736,0.001184308,0.113406,0.008575084,0.4653957,0.3700503],"study_design_scores_gemma":[0.00006813397,0.00005866048,0.01528826,0.001231736,0.000222334,0.001110944,0.0002729458,0.0004688376,0.00910223,0.002196707,0.9699042,0.00007491492],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.007208652,0.01558021,0.01611831,0.0007515872,0.00087114,0.000428574,0.9364567,0.007435163,0.01514962],"genre_scores_gemma":[0.006374523,0.01011878,0.04102429,0.0005545453,0.0001922164,0.0005335024,0.9353644,0.001720798,0.004116963],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.02302446,"threshold_uncertainty_score":0.07702452,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01668037882140654,"score_gpt":0.2301297035441332,"score_spread":0.2134493247227267,"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."}}