{"id":"W2009070690","doi":"10.5731/pdajpst.2014.01023","title":"Cataloguing the Taxonomic Origins of Sequences from a Heterogeneous Sample Using Phylogenomics: Applications in Adventitious Agent Detection","year":2014,"lang":"en","type":"article","venue":"PDA Journal of Pharmaceutical Science and Technology","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"Sanofi (Canada)","funders":"","keywords":"Contig; Phylogenomics; Computer science; Metagenomics; Taxonomic rank; Sequence (biology); Software; Classifier (UML); Artificial intelligence; Data mining; Biology; Phylogenetic tree; Genome; Genetics; Taxon","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":[0.001889052,0.0007732434,0.0008440249,0.00187566,0.0005696933,0.0008471915,0.0006947973,0.000543212,0.0006472613],"category_scores_gemma":[0.003210563,0.0005331668,0.0005991591,0.001572723,0.0004523422,0.0008511571,0.0007319221,0.0005504463,0.0005001558],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004533629,"about_ca_system_score_gemma":0.0006506178,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001949627,"about_ca_topic_score_gemma":0.00286883,"domain_scores_codex":[0.9994578,0.00009452157,0.00005782894,0.000244676,0.0001094,0.000035832],"domain_scores_gemma":[0.9988019,0.0005291795,0.0002468781,0.0001429613,0.0002036864,0.00007531274],"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":[0.001141074,0.0001297113,0.06175372,0.0005119099,0.0001187998,0.0002896071,0.000921679,0.01178182,0.7752647,0.0007723508,0.0004493021,0.1468653],"study_design_scores_gemma":[0.00005256299,0.0006206742,0.07803288,0.0001196627,0.000199823,0.0009556006,0.000464361,0.1999311,0.7087055,0.001787611,0.00899738,0.0001328135],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5351394,0.0008524912,0.4549989,0.000138323,0.0000152886,0.0001824682,0.003037845,0.005038512,0.0005967966],"genre_scores_gemma":[0.265636,0.0003451175,0.7296242,0.00006395885,0.000006385131,0.0001637336,0.003435484,0.0002716454,0.0004534536],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.001949627,"threshold_uncertainty_score":0.009990394,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02237278395954507,"score_gpt":0.2949805395617308,"score_spread":0.2726077556021858,"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."}}