{"id":"W2120895683","doi":"10.1186/1471-2105-12-8","title":"Flexible taxonomic assignment of ambiguous sequencing reads","year":2011,"lang":"en","type":"article","venue":"BMC Bioinformatics","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":32,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Institute of Genetics","keywords":"Metagenomics; Computer science; Set (abstract data type); False positive paradox; Data mining; Computational biology; Matching (statistics); Phylogenetic tree; Biology; Information retrieval; Artificial intelligence; Statistics; Genetics; Mathematics; Gene","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":[0.005761091,0.001300417,0.001076488,0.002148312,0.001175675,0.001507253,0.002066625,0.001588393,0.001438523],"category_scores_gemma":[0.02148804,0.0005175457,0.00101074,0.002017365,0.001164305,0.001379831,0.002364662,0.00172661,0.00119847],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007178159,"about_ca_system_score_gemma":0.001235593,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001057443,"about_ca_topic_score_gemma":0.00130763,"domain_scores_codex":[0.9947365,0.001205076,0.0004677979,0.001842775,0.001499241,0.0002485756],"domain_scores_gemma":[0.9848993,0.005944697,0.002329451,0.003200198,0.003180934,0.0004454082],"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":[0.001723578,0.0005307359,0.04813216,0.001288056,0.0003197872,0.0008150429,0.001561234,0.1293019,0.2903928,0.007897065,0.004497025,0.5135407],"study_design_scores_gemma":[0.000189865,0.0004809322,0.02130721,0.000228697,0.0001476706,0.001247141,0.0004656751,0.7553355,0.1800579,0.02814441,0.0122524,0.0001425773],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1432374,0.0004330844,0.8516765,0.0001144172,0.0000750722,0.000226956,0.0005279728,0.002825962,0.0008826702],"genre_scores_gemma":[0.3171127,0.0001264468,0.6787603,0.0002320697,0.00005463612,0.0002677894,0.00229344,0.0004972609,0.0006552862],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005761091,"threshold_uncertainty_score":0.03046793,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.062467156920103,"score_gpt":0.231665825668583,"score_spread":0.1691986687484799,"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."}}