{"id":"W4391512208","doi":"10.1038/s41467-024-45198-y","title":"The impacts of active and self-supervised learning on efficient annotation of single-cell expression data","year":2024,"lang":"en","type":"article","venue":"Nature Communications","topic":"Single-cell and spatial transcriptomics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; Sinai Health System; Lunenfeld-Tanenbaum Research Institute; Ontario Institute for Cancer Research; University of Toronto","funders":"Canadian Institutes of Health Research; Natural Sciences and Engineering Research Council of Canada; Canada Research Chairs; Princess Margaret Cancer Foundation","keywords":"Annotation; Computer science; Benchmarking; Classifier (UML); Machine learning; Artificial intelligence; Heuristic; Supervised learning; Active learning (machine learning); Crowdsourcing; Similarity (geometry); Data mining; World Wide Web","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":[0.01678341,0.001504684,0.001468951,0.001256625,0.001056521,0.002808511,0.003031329,0.002273124,0.001031236],"category_scores_gemma":[0.03697699,0.0006337189,0.001073183,0.001447168,0.002626257,0.004680023,0.002537295,0.00273633,0.0009614945],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001487543,"about_ca_system_score_gemma":0.001405431,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003477545,"about_ca_topic_score_gemma":0.004996032,"domain_scores_codex":[0.9924123,0.003690511,0.0003308744,0.001852307,0.00139848,0.0003154974],"domain_scores_gemma":[0.962195,0.02654914,0.001573565,0.006252109,0.002736852,0.000693311],"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":[0.001718753,0.0007378286,0.01797833,0.001141832,0.0005427819,0.0001980672,0.0009332562,0.4314953,0.04687666,0.01680221,0.00984518,0.4717299],"study_design_scores_gemma":[0.00005314027,0.0001710002,0.002407139,0.00006179577,0.00004359185,0.00008682696,0.0001238379,0.9552698,0.02260219,0.01637345,0.002762782,0.00004441349],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1567194,0.002889457,0.8283303,0.001735073,0.0002999591,0.0001773121,0.0007183728,0.00508961,0.004040441],"genre_scores_gemma":[0.5430213,0.0009155979,0.4479091,0.0008524989,0.0001618138,0.0003585005,0.002681446,0.001240522,0.002859172],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01678341,"threshold_uncertainty_score":0.08876026,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02241204501607454,"score_gpt":0.2859901883827774,"score_spread":0.2635781433667029,"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."}}