{"id":"W6958184062","doi":"10.6084/m9.figshare.16620928.v1","title":"Additional file 2 of PathFams: statistical detection of pathogen-associated protein domains","year":2021,"lang":"en","type":"article","venue":"Figshare","topic":"Machine Learning in Bioinformatics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Wilfrid Laurier University; University of Waterloo","funders":"","keywords":"Domain (mathematical analysis); Statistical analysis; Pattern recognition (psychology); Protein detection; Table (database)","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":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.00002186449,0.00007304007,0.0001004228,0.00001641295,0.00002748778,0.000005444097,0.00006956601,0.0001182561,0.9121847],"category_scores_gemma":[0.008043895,0.00007607705,0.00005642928,0.00008081085,0.00001159759,0.000002350873,0.00008048825,0.00007962085,0.0001296446],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000009363638,"about_ca_system_score_gemma":0.0001334046,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":6.651571e-7,"about_ca_topic_score_gemma":0.00001224944,"domain_scores_codex":[0.9993944,0.00005508945,0.0002003703,0.0001093881,0.0001392147,0.0001015519],"domain_scores_gemma":[0.9992621,0.0001302964,0.0001852621,0.0001517394,0.0002373093,0.00003331652],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00000850894,0.00004853319,0.000002774225,0.00010756,0.00002464076,0.000003516694,0.00001007383,0.000007507231,0.06584413,0.000004051816,0.9328675,0.001071239],"study_design_scores_gemma":[0.0002589194,0.0002161688,0.00434613,0.0007861041,0.000006754239,0.00001353317,0.00003738261,0.000744068,0.1249186,0.00003590569,0.8684792,0.0001572212],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.001104992,0.00001706363,0.00004955885,0.000003385996,0.000006941481,0.00008231273,0.9973322,0.000007441666,0.00139607],"genre_scores_gemma":[0.03088105,4.087166e-7,0.002938161,0.00002588745,0.00003644103,0.0001685132,0.965483,0.00001002088,0.000456534],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.912055,"threshold_uncertainty_score":0.9629875,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.006946089948129705,"score_gpt":0.2201791178083597,"score_spread":0.21323302786023,"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."}}