{"id":"W4402841129","doi":"10.2139/ssrn.4965919","title":"A Computational Framework for Defining and Validating Reproducible Phenotyping Algorithms of 313 Diseases in the UK Biobank","year":2024,"lang":"en","type":"preprint","venue":"SSRN Electronic Journal","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"Institute of Cancer Research","funders":"Medical Research Council","keywords":"Biobank; Computer science; Computational biology; Algorithm; Biology; Bioinformatics","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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.03717437,0.001235291,0.001727292,0.006630366,0.002444896,0.0105209,0.005120225,0.002871851,0.002339553],"category_scores_gemma":[0.09837697,0.00146953,0.003919052,0.005112951,0.002633678,0.007285945,0.007438082,0.003291548,0.0008012776],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00521115,"about_ca_system_score_gemma":0.01292427,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02137796,"about_ca_topic_score_gemma":0.03040792,"domain_scores_codex":[0.9767181,0.008500464,0.00490071,0.003857391,0.00514574,0.000877561],"domain_scores_gemma":[0.9323561,0.04177221,0.004289741,0.01320671,0.007012467,0.001362757],"domain_codex":null,"domain_gemma":"reproducibility","domain_candidate":"reproducibility","domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0006668404,0.0008091771,0.01588162,0.001739176,0.001198432,0.001024293,0.001587971,0.4658808,0.007552663,0.2843519,0.01908637,0.2002207],"study_design_scores_gemma":[0.000123182,0.000107778,0.001271931,0.00032451,0.0002073472,0.0001713831,0.0002189372,0.8842733,0.004186606,0.09728209,0.01177969,0.00005333355],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.008848085,0.0001425525,0.9813176,0.0007938542,0.00003330051,0.0005920681,0.002176401,0.005327196,0.0007689461],"genre_scores_gemma":[0.07783584,0.0001073133,0.9140823,0.0002081659,0.00002158184,0.0005320089,0.006603944,0.0003527282,0.0002561019],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9628257,"threshold_uncertainty_score":0.1965991,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01860545006647126,"score_gpt":0.316032856469037,"score_spread":0.2974274064025658,"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."}}