{"id":"W4406715995","doi":"10.1101/2025.01.20.633936","title":"PDP-Miner: an AI/ML tool to detect prophage tail proteins with depolymerase domains across thousands of bacterial genomes","year":2025,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Machine Learning in Bioinformatics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal; Université Laval","funders":"Canadian Institutes of Health Research; Ontario Genomics; Genome Canada","keywords":"Prophage; Bacterial genome size; Genome; Computational biology; Biology; Genetics; Bacteriophage; Gene; Escherichia coli","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":[0.002349422,0.003470648,0.001390425,0.003281283,0.001011847,0.001952377,0.003266209,0.001424685,0.01333538],"category_scores_gemma":[0.007027714,0.001669179,0.001864817,0.001669841,0.0006269179,0.00212574,0.00272749,0.002331173,0.00899665],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005332835,"about_ca_system_score_gemma":0.001361097,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00108322,"about_ca_topic_score_gemma":0.001601043,"domain_scores_codex":[0.998745,0.0001818923,0.0001393274,0.0005119099,0.0003422997,0.00007962049],"domain_scores_gemma":[0.9973573,0.001641379,0.0003504158,0.0002523173,0.0002233679,0.0001751555],"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":[0.003322199,0.0006802253,0.03084753,0.005164875,0.001259416,0.001973413,0.0008859566,0.03317691,0.1109387,0.005017679,0.3781674,0.4285656],"study_design_scores_gemma":[0.001125732,0.0005623986,0.01072099,0.000354893,0.0003735021,0.001217678,0.0003243011,0.6808037,0.1307823,0.01928743,0.1541573,0.0002898996],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"software","genre_gemma":"methods","genre_scores_codex":[0.0227638,0.0008008899,0.2527186,0.0005446511,0.0001842702,0.0003848811,0.03855483,0.6815588,0.002489314],"genre_scores_gemma":[0.1195145,0.0006154479,0.7335121,0.000763773,0.0001459531,0.001333401,0.09840913,0.04100113,0.00470461],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.01333538,"threshold_uncertainty_score":0.04461128,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.005373435462376582,"score_gpt":0.2371197319489669,"score_spread":0.2317462964865903,"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."}}