{"id":"W4406825074","doi":"10.1101/2025.01.23.634626","title":"CANDI: self-supervised, confidence-aware denoising imputation of genomic data","year":2025,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Gene expression and cancer classification","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"","keywords":"Imputation (statistics); Computer science; Noise reduction; Artificial intelligence; Confidence interval; Statistics; Machine learning; Mathematics; Missing data","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":[0.008534452,0.001020358,0.001770301,0.001130715,0.0006355289,0.001384001,0.005273069,0.002235348,0.002094805],"category_scores_gemma":[0.02008217,0.0008445342,0.001342664,0.001450164,0.001384532,0.00125509,0.002419175,0.004575036,0.001444694],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001253129,"about_ca_system_score_gemma":0.002374049,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003836387,"about_ca_topic_score_gemma":0.006392356,"domain_scores_codex":[0.9976908,0.0007993945,0.00008707624,0.0006900814,0.0005891381,0.0001436017],"domain_scores_gemma":[0.9917907,0.003846596,0.0006529083,0.00208423,0.001265063,0.0003605768],"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":[0.0008475168,0.0005351183,0.01268354,0.0003817939,0.0006135358,0.0002436737,0.0002270269,0.6352001,0.01478782,0.01950356,0.03560143,0.2793748],"study_design_scores_gemma":[0.00002050038,0.00002359473,0.0004366659,0.000009195324,0.00001064713,0.00002520599,0.000005146793,0.9885794,0.003123379,0.006769437,0.0009825581,0.00001439826],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0080155,0.0002553262,0.9861389,0.0003621443,0.00006839668,0.00006008029,0.0006865505,0.003897875,0.0005152272],"genre_scores_gemma":[0.1795727,0.0001983691,0.8065197,0.0008875589,0.0002003292,0.0004102514,0.006527643,0.0009016162,0.004781824],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.008534452,"threshold_uncertainty_score":0.04513502,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02096103433872334,"score_gpt":0.258679237263246,"score_spread":0.2377182029245227,"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."}}