{"id":"W4386258894","doi":"10.1016/j.csbj.2023.08.025","title":"MetaPep: A core peptide database for faster human gut metaproteomics database searches","year":2023,"lang":"en","type":"article","venue":"Computational and Structural Biotechnology Journal","topic":"Advanced Proteomics Techniques and Applications","field":"Chemistry","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Metaproteomics; Database search engine; Database; Sequence database; Computer science; Computational biology; Metagenomics; Biology; Search engine; Information retrieval","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00371191,0.002330913,0.001947291,0.004423896,0.001110529,0.002954065,0.002711368,0.001279629,0.01354873],"category_scores_gemma":[0.006559738,0.001245906,0.001627815,0.004968561,0.0003149504,0.004139439,0.004181295,0.002136226,0.01208869],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005641443,"about_ca_system_score_gemma":0.001784927,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0008671155,"about_ca_topic_score_gemma":0.0009603382,"domain_scores_codex":[0.9985647,0.0002226916,0.0002406069,0.0004225961,0.0003966048,0.0001528061],"domain_scores_gemma":[0.9983852,0.0004068641,0.000211095,0.0004585923,0.0003422756,0.0001958805],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.007760567,0.0009171445,0.01182417,0.006316738,0.001373605,0.002626561,0.001016064,0.004543244,0.2604167,0.01143711,0.2281989,0.4635691],"study_design_scores_gemma":[0.002254501,0.001089178,0.02298219,0.000864293,0.0007498185,0.00465222,0.000811179,0.1022269,0.26684,0.03350877,0.5634387,0.0005823347],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.06013811,0.00647566,0.4580441,0.001064917,0.0006926965,0.001435137,0.2661706,0.196282,0.00969669],"genre_scores_gemma":[0.06916332,0.002554202,0.627083,0.0005705205,0.0001211621,0.001564629,0.288559,0.008341437,0.002042697],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01354873,"threshold_uncertainty_score":0.04532504,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05752189728448905,"score_gpt":0.3402681764971878,"score_spread":0.2827462792126987,"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."}}