{"id":"W4394675060","doi":"10.1093/bioinformatics/btae192","title":"VirusPredictor: XGBoost-based software to predict virus-related sequences in human data","year":2024,"lang":"en","type":"article","venue":"Bioinformatics","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"Centre Hospitalier Universitaire Sainte-Justine","funders":"National Institute of Allergy and Infectious Diseases; National Institutes of Health; Open Medicine Foundation; U.S. Department of Defense","keywords":"Python (programming language); Contig; Genome; Metagenomics; Biology; Sanger sequencing; Computational biology; Software; Genetics; DNA sequencing; Genomics; Human genome; Computer science; 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":[0.001672092,0.001703797,0.000888413,0.001427835,0.0004779652,0.0007304971,0.001689751,0.001207248,0.008169247],"category_scores_gemma":[0.003099554,0.0009246044,0.001458641,0.001024011,0.000338104,0.0007460208,0.0007178762,0.001646276,0.004158637],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006441748,"about_ca_system_score_gemma":0.001240376,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005178027,"about_ca_topic_score_gemma":0.00632742,"domain_scores_codex":[0.9994331,0.0001419159,0.0000503107,0.000196291,0.0001294004,0.00004895417],"domain_scores_gemma":[0.9992812,0.0004688684,0.00006757786,0.0000584512,0.0000874234,0.00003651063],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002372488,0.0006118831,0.02253455,0.002550952,0.001622899,0.0006022916,0.0004751324,0.2680146,0.02828529,0.004701976,0.1746141,0.4936138],"study_design_scores_gemma":[0.0001161271,0.0001227336,0.002555336,0.00005366365,0.00005257393,0.0001008567,0.00002964148,0.9732481,0.008764538,0.002686706,0.01223489,0.00003497663],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"software","genre_scores_codex":[0.06578057,0.001957458,0.5753656,0.0004434877,0.0004062424,0.0006663376,0.02439267,0.32764,0.003347598],"genre_scores_gemma":[0.167483,0.0007799757,0.7728612,0.0006970678,0.0001244594,0.00147063,0.0399255,0.01179952,0.004858661],"genre_candidate":"software","genre_consensus":null,"teacher_disagreement_score":0.008169247,"threshold_uncertainty_score":0.02732879,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02375274263784608,"score_gpt":0.274237705772119,"score_spread":0.2504849631342729,"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."}}