{"id":"W2979549191","doi":"10.1093/gigascience/giz119","title":"PIRATE: A fast and scalable pangenomics toolbox for clustering diverged orthologues in bacteria","year":2019,"lang":"en","type":"article","venue":"GigaScience","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":272,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Biotechnology and Biological Sciences Research Council; Medical Research Council; Medical Research Council Canada; Sight Research UK","keywords":"Genome; Cluster analysis; Biology; Scalability; Bacterial genome size; Toolbox; Computational biology; Gene; Computer science; Data mining; Genetics; Database; Artificial intelligence","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.000185881,0.00009093047,0.0001103586,0.00002605806,0.000065217,0.00003196315,0.0001374122,0.00004997333,0.000003615833],"category_scores_gemma":[0.00002564051,0.00008520734,0.00002520856,0.00005605402,0.00007681472,0.000001937743,0.0001762579,0.00002653532,0.000004220256],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000007634902,"about_ca_system_score_gemma":0.00002833821,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001873503,"about_ca_topic_score_gemma":0.0001469911,"domain_scores_codex":[0.9992889,0.00001227289,0.0001139078,0.0003193119,0.00004032445,0.000225342],"domain_scores_gemma":[0.9997289,0.000009724332,0.00003580029,0.0001608617,0.00002888264,0.00003580632],"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":[0.00002842929,0.000009591643,0.07320098,0.00001448531,0.000005070556,3.494174e-7,0.0001354783,0.0001715761,0.9241385,0.00002243104,0.00003375803,0.002239396],"study_design_scores_gemma":[0.001793232,0.0009353283,0.8213997,0.000033184,0.0000152814,0.0000217147,0.0007270153,0.002827354,0.1474044,0.000217967,0.02399604,0.0006287643],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9981968,0.0004423963,0.0005372333,0.00004886524,0.0001902233,0.0002338808,0.00002417626,0.000002012271,0.0003243934],"genre_scores_gemma":[0.9960448,0.0001790161,0.003192522,0.0001483119,0.00003962078,0.00002300939,0.000008576066,0.000007561633,0.0003566105],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7767341,"threshold_uncertainty_score":0.3474655,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01159327289252637,"score_gpt":0.2236301625589356,"score_spread":0.2120368896664092,"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."}}