{"id":"W4386798604","doi":"10.1038/s41597-023-02502-7","title":"Whole genomes from bacteria collected at diagnostic units around the world 2020","year":2023,"lang":"en","type":"article","venue":"Scientific Data","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"Centre Hospitalier Universitaire de Sherbrooke","funders":"Horizon 2020 Framework Programme; Novo Nordisk Fonden; European Commission","keywords":"Genome; Bacteria; Biology; Computational biology; Evolutionary biology; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001386145,0.0009834433,0.0008747431,0.005998345,0.0007858164,0.001275311,0.000572796,0.001011448,0.00417343],"category_scores_gemma":[0.003841858,0.0004281295,0.0009354647,0.009467348,0.0003712419,0.0007396297,0.001731937,0.0008384254,0.003571822],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009369445,"about_ca_system_score_gemma":0.002777348,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01735809,"about_ca_topic_score_gemma":0.0252621,"domain_scores_codex":[0.9989084,0.0001284972,0.000137503,0.0002821127,0.0003540255,0.000189429],"domain_scores_gemma":[0.9978143,0.0003618252,0.0004973679,0.0002754591,0.0005771857,0.0004738758],"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":[0.005619759,0.0007255437,0.2575567,0.01297202,0.001493973,0.001672165,0.001701181,0.005963343,0.1206582,0.004164306,0.397037,0.1904358],"study_design_scores_gemma":[0.0002245702,0.0003580082,0.5355191,0.001161658,0.0004570152,0.001162764,0.001432077,0.001294402,0.01368017,0.001862533,0.4427057,0.0001420862],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.03724597,0.00239213,0.001130353,0.0003137938,0.00005760094,0.00007735583,0.9563438,0.0002980782,0.002140927],"genre_scores_gemma":[0.01595177,0.0008688486,0.003110409,0.000121007,0.00001479988,0.00007146644,0.9794901,0.0000288117,0.0003428028],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.01735809,"threshold_uncertainty_score":0.03451413,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04050191828275441,"score_gpt":0.2594238348073313,"score_spread":0.2189219165245769,"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."}}