{"id":"W4293066864","doi":"10.1093/nar/gkac689","title":"Metagenomics versus total RNA sequencing: most accurate data-processing tools, microbial identification accuracy and perspectives for ecological assessments","year":2022,"lang":"en","type":"article","venue":"Nucleic Acids Research","topic":"Microbial Community Ecology and Physiology","field":"Environmental Science","cited_by":47,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph","funders":"Canada First Research Excellence Fund; University of Guelph","keywords":"Biology; Metagenomics; Identification (biology); Computational biology; Microbial ecology; Ecology; DNA sequencing; Data science; Genetics; Bacteria; Gene; Computer science","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":["sts","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.002562046,0.0001385862,0.0001975543,0.00007752575,0.002437219,0.0001846392,0.001222343,0.0001051184,0.004205976],"category_scores_gemma":[0.0007642682,0.0001427621,0.00003510412,0.000359419,0.0006075226,0.0006834714,0.003954277,0.000847951,0.00007190997],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008122399,"about_ca_system_score_gemma":0.0001538492,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001627158,"about_ca_topic_score_gemma":0.0001656232,"domain_scores_codex":[0.9974609,0.0008741273,0.000256925,0.0006315137,0.0002613963,0.0005151539],"domain_scores_gemma":[0.9981868,0.0009447659,0.0001148825,0.0006114304,0.00005262401,0.00008955801],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"not_applicable","study_design_scores_codex":[0.001228757,0.0003938393,0.0003696834,0.00002841197,0.00007649726,0.00001011777,0.00268668,0.0009683851,0.9542971,0.0005431183,0.01431701,0.02508041],"study_design_scores_gemma":[0.01849603,0.01052491,0.2939592,0.00003557876,0.000454601,0.0003479024,0.1428621,0.1914705,0.02110872,0.01240553,0.3045107,0.003824183],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9966514,0.0000810353,0.0001773869,0.0005885657,0.0001854272,0.0005986181,0.0003124935,0.00002969605,0.001375446],"genre_scores_gemma":[0.9972862,0.0000573639,0.001444409,0.00008436508,0.00006238942,0.00008987659,0.0005378337,0.00001908942,0.0004184661],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9331884,"threshold_uncertainty_score":0.9988615,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2591727983652958,"score_gpt":0.4171740669071816,"score_spread":0.1580012685418858,"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."}}