{"id":"W3047701039","doi":"10.1101/2020.08.07.241729","title":"Petabase-scale sequence alignment catalyses viral discovery","year":2020,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Bacteriophages and microbial interactions","field":"Environmental Science","cited_by":57,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"Klaus Tschira Stiftung; University of British Columbia; Russian Science Foundation; Ministerio de Economía y Competitividad; Saint Petersburg State University; Agence Nationale de la Recherche","keywords":"Biology; Computational biology; Sequence (biology); RNA polymerase; RNA; Polymerase; RNA virus; Virology; Nucleic acid; Anticipation (artificial intelligence); Gene; Genetics; Computer science; Artificial intelligence","routes":{"ca_aff":true,"ca_fund":true,"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":["metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0001720679,0.0005713787,0.0004813,0.00006182947,0.0002185501,0.0003993239,0.0007330301,0.0002700312,0.0008896232],"category_scores_gemma":[0.00004604249,0.0005947683,0.0002605403,0.0002653658,0.0002629135,0.0006421361,0.001663606,0.0007409761,0.001365693],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007335834,"about_ca_system_score_gemma":0.0001359426,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001108345,"about_ca_topic_score_gemma":0.00003815941,"domain_scores_codex":[0.997234,0.00009199794,0.0004942007,0.001244486,0.0003639267,0.0005713241],"domain_scores_gemma":[0.9981785,0.0000298945,0.0003403465,0.001076171,0.00003303249,0.0003420328],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00001695821,0.000117901,0.001767804,0.0000689151,0.00005905769,0.00007100985,0.00001856145,0.0001226992,0.9959962,0.000034261,0.001724402,0.000002208024],"study_design_scores_gemma":[0.0002525974,0.0000576502,0.113304,0.0002050569,0.0001738864,8.358819e-8,0.000006330503,0.0002061512,0.8575628,0.000002636091,0.02720096,0.001027875],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9935896,0.0001471867,0.001925152,0.0008218345,0.001473002,0.0006000512,0.001035446,0.000303983,0.0001037878],"genre_scores_gemma":[0.9945522,0.0001428642,0.00392649,0.0007197617,0.0003470075,0.0001400902,0.000002242078,0.0001003128,0.00006903873],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1384334,"threshold_uncertainty_score":0.9996504,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01987457877817397,"score_gpt":0.235009707410126,"score_spread":0.2151351286319521,"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."}}