{"id":"W6976907603","doi":"10.60692/zgkw1-6av98","title":"MiDAS 4: A global catalogue of full-length 16S rRNA gene sequences and taxonomy for studies of bacterial communities in wastewater treatment plants","year":2022,"lang":"en","type":"article","venue":"Greater South Information System","topic":"Wastewater Treatment and Nitrogen Removal","field":"Environmental Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"16S ribosomal RNA; Taxonomy (biology); Ribosomal RNA; Sewage treatment; Activated sludge; Amplicon sequencing; Gene sequence; 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":[0.002476662,0.00260338,0.001465156,0.00934902,0.001098659,0.001578058,0.00142561,0.001104414,0.004324824],"category_scores_gemma":[0.005035122,0.0007674108,0.001468596,0.008304496,0.0003202088,0.001805614,0.00233105,0.001122851,0.006327426],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007707084,"about_ca_system_score_gemma":0.003109252,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003084145,"about_ca_topic_score_gemma":0.006371653,"domain_scores_codex":[0.9981871,0.0002628388,0.0003158909,0.0004968019,0.0005262597,0.0002109626],"domain_scores_gemma":[0.9970883,0.0004155576,0.001060444,0.0003975854,0.0007389956,0.0002992113],"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":[0.00431755,0.0005316342,0.1017634,0.0128915,0.0009858297,0.0008222914,0.002202808,0.005875811,0.3781224,0.003584982,0.1272051,0.3616967],"study_design_scores_gemma":[0.0002624155,0.0008597377,0.2301641,0.001597825,0.0007637534,0.001166253,0.0008165358,0.009066272,0.09088258,0.003827023,0.660291,0.0003024707],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.1203385,0.005897612,0.04741199,0.0001922184,0.0001696527,0.0005171078,0.8052284,0.01217581,0.008068663],"genre_scores_gemma":[0.0418823,0.001293625,0.08209345,0.0001054488,0.00003957955,0.0006747692,0.8709584,0.001128571,0.001823869],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.00934902,"threshold_uncertainty_score":0.01446795,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0671128673543027,"score_gpt":0.2298036699372068,"score_spread":0.1626908025829041,"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."}}