{"id":"W4415779768","doi":"10.1101/2025.11.01.686014","title":"Oyster: a neural network for modelling genomic sequences that enables exact position-specific <i>k</i> -mer contributions","year":2025,"lang":"","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Genomics and Chromatin Dynamics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Interpretability; Function (biology); DNA sequencing; Histone; Convolutional neural network; ENCODE; Genomics; Artificial neural network","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":[0.0007132541,0.0009093527,0.0003934565,0.0003887641,0.0002442861,0.0008190103,0.001603396,0.00103899,0.00226137],"category_scores_gemma":[0.002450137,0.0005069243,0.0006415662,0.0004108501,0.0005174838,0.001136191,0.001305698,0.001579462,0.0005574105],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008862583,"about_ca_system_score_gemma":0.0009501749,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01331817,"about_ca_topic_score_gemma":0.01711491,"domain_scores_codex":[0.9998523,0.00002491446,0.000009174926,0.00005563327,0.00003522003,0.00002282103],"domain_scores_gemma":[0.9996136,0.0002138257,0.00003587797,0.00004820868,0.00006398026,0.00002455196],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001533006,0.00005434452,0.001947882,0.00009330386,0.00006585841,0.00008767428,0.00005656758,0.9179376,0.008563379,0.005114054,0.002214424,0.06371158],"study_design_scores_gemma":[0.000002782058,0.000007504718,0.00008271223,0.000003356439,0.000002782902,0.00000517353,0.000002892775,0.9971713,0.0009926031,0.001477814,0.0002484835,0.000002608678],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"software","genre_scores_codex":[0.05588655,0.0002652535,0.9364985,0.0002427683,0.00005370907,0.00004653047,0.000678259,0.004727271,0.001601294],"genre_scores_gemma":[0.6032355,0.0005118301,0.3847269,0.0003809457,0.00004035092,0.0002503016,0.002692245,0.0007248002,0.007437155],"genre_candidate":"software","genre_consensus":null,"teacher_disagreement_score":0.01331817,"threshold_uncertainty_score":0.02648133,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01299530025112487,"score_gpt":0.2154931762356102,"score_spread":0.2024978759844853,"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."}}