{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002558379,0.00229162,0.001595133,0.003447274,0.001034921,0.001954521,0.002997886,0.000975034,0.01332651],"category_scores_gemma":[0.005624806,0.001350575,0.002216031,0.002187748,0.000874861,0.002682997,0.004032892,0.00212153,0.007075069],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008324034,"about_ca_system_score_gemma":0.001696435,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00174964,"about_ca_topic_score_gemma":0.002786989,"domain_scores_codex":[0.9988354,0.0001786334,0.00010168,0.0003973802,0.0003806321,0.0001063115],"domain_scores_gemma":[0.9984686,0.0005679412,0.0002302406,0.0003066169,0.0002345713,0.0001919841],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.003634345,0.0004880088,0.01647842,0.0045868,0.001479557,0.00132934,0.002415272,0.05026626,0.2146356,0.01828179,0.1359376,0.550467],"study_design_scores_gemma":[0.0009849868,0.000600994,0.01846936,0.0005168921,0.0003398921,0.001681825,0.0007079144,0.5942883,0.1217682,0.06576679,0.1943312,0.0005436406],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"software","genre_scores_codex":[0.01420886,0.0006654306,0.7496327,0.0002476706,0.00008612047,0.0002878527,0.01135046,0.2216228,0.001898164],"genre_scores_gemma":[0.06035657,0.0006880153,0.8900902,0.0002644074,0.00005375965,0.001156316,0.02482635,0.02079701,0.001767292],"genre_candidate":"software","genre_consensus":null,"teacher_disagreement_score":0.01332651,"threshold_uncertainty_score":0.04458165,"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."}}