{"id":"W4308678857","doi":"10.1093/bib/bbac443","title":"High-resolution shotgun metagenomics: the more data, the better?","year":2022,"lang":"en","type":"article","venue":"Briefings in Bioinformatics","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":28,"is_retracted":false,"has_abstract":true,"ca_institutions":"National Research Council Canada","funders":"","keywords":"Metagenomics; Shotgun sequencing; Deep sequencing; Workflow; DNA sequencing; Computer science; Sample (material); Shotgun; Data mining; Computational biology; Biology; Genome; Database; Genetics; 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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.01118134,0.00139769,0.001883658,0.001688418,0.001150244,0.006296552,0.00166643,0.002965787,0.001776534],"category_scores_gemma":[0.01320799,0.0004911309,0.001278923,0.004339948,0.001789993,0.007386962,0.002085505,0.003871409,0.0008521099],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001611429,"about_ca_system_score_gemma":0.001835999,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003645505,"about_ca_topic_score_gemma":0.01098956,"domain_scores_codex":[0.9953667,0.001872941,0.0002260285,0.001044566,0.001193765,0.0002959243],"domain_scores_gemma":[0.9908432,0.004504323,0.0007086495,0.001315269,0.002000441,0.0006281419],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.002271919,0.0005337666,0.116973,0.0114073,0.001994577,0.00219339,0.003969587,0.03009203,0.2304492,0.03442299,0.05452764,0.5111645],"study_design_scores_gemma":[0.0002545579,0.0008093715,0.09295508,0.003721213,0.001392121,0.002300234,0.01296545,0.0892301,0.1750896,0.223159,0.3971489,0.0009743854],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.4707448,0.09628306,0.3187006,0.0737975,0.002756458,0.0004510882,0.02209142,0.003265403,0.01190962],"genre_scores_gemma":[0.5519632,0.02978281,0.3713715,0.01328917,0.001178959,0.0004112403,0.02833725,0.0009131521,0.002752563],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9888186,"threshold_uncertainty_score":0.05913329,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01890786310678922,"score_gpt":0.2353498665392683,"score_spread":0.2164420034324791,"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."}}