{"id":"W4283018586","doi":"10.1016/j.ygeno.2022.110414","title":"VirusTaxo: Taxonomic classification of viruses from the genome sequence using k-mer enrichment","year":2022,"lang":"en","type":"article","venue":"Genomics","topic":"Bacteriophages and microbial interactions","field":"Environmental Science","cited_by":27,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"University of British Columbia","keywords":"Metagenomics; Biology; Virus classification; Taxonomy (biology); Contig; DNA sequencing; Genome; Biological classification; Taxonomic rank; Computational biology; Evolutionary biology; Genetics; Taxon; Gene; Ecology","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001251983,0.002219855,0.001527536,0.004377068,0.001403767,0.002114195,0.00114738,0.00103192,0.003404275],"category_scores_gemma":[0.002942401,0.0005898972,0.00234069,0.001972813,0.0004664042,0.001735898,0.001725814,0.001336055,0.003262616],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007658745,"about_ca_system_score_gemma":0.001428146,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003294792,"about_ca_topic_score_gemma":0.005477353,"domain_scores_codex":[0.9991597,0.00009626048,0.00008284935,0.0003240699,0.0002085512,0.0001284719],"domain_scores_gemma":[0.99923,0.0002459929,0.0001342983,0.00008929797,0.0001921824,0.0001083232],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.003748919,0.001016643,0.09254023,0.004489291,0.001696782,0.0007161032,0.001555788,0.01099653,0.4747999,0.00363207,0.05209826,0.3527095],"study_design_scores_gemma":[0.0003887691,0.001645376,0.07628398,0.000588534,0.0007924135,0.001821641,0.00170996,0.5027553,0.2710682,0.01406835,0.1282928,0.0005846684],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.4818687,0.004152803,0.3237718,0.0007950988,0.0009061602,0.002130332,0.102888,0.07435872,0.009128442],"genre_scores_gemma":[0.238708,0.001420899,0.6013229,0.0005177242,0.0001310748,0.001499357,0.1452986,0.00494746,0.006154008],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.004377068,"threshold_uncertainty_score":0.01138848,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07357802934634133,"score_gpt":0.2723540530831424,"score_spread":0.198776023736801,"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."}}