{"id":"W4379276041","doi":"10.1093/bioadv/vbad068","title":"DataCurator.jl: efficient, portable and reproducible validation, curation and transformation of large heterogeneous datasets using human-readable recipes compiled into machine-verifiable templates","year":2023,"lang":"en","type":"article","venue":"Bioinformatics Advances","topic":"Scientific Computing and Data Management","field":"Decision Sciences","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia; Simon Fraser University","funders":"","keywords":"Computer science; Executable; Scalability; Python (programming language); Preprocessor; Workflow; Data mining; Verifiable secret sharing; Data curation; Database; Programming language","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":["open_science"],"consensus_categories":[],"category_scores_codex":[0.01245636,0.003507793,0.002660649,0.003507976,0.00175021,0.00596305,0.005597734,0.001691314,0.05162053],"category_scores_gemma":[0.03450952,0.004077805,0.003769364,0.002629879,0.002894021,0.004989835,0.01013978,0.006130868,0.08045998],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00157964,"about_ca_system_score_gemma":0.005893007,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003436417,"about_ca_topic_score_gemma":0.003879301,"domain_scores_codex":[0.9913855,0.001358859,0.001000771,0.002297995,0.003323874,0.0006330079],"domain_scores_gemma":[0.9788825,0.007056053,0.001726542,0.007499284,0.003902416,0.0009332133],"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":[0.001048883,0.0001678982,0.005118283,0.003222468,0.0004845072,0.0006346257,0.001455697,0.005220192,0.03225213,0.01170167,0.8247947,0.113899],"study_design_scores_gemma":[0.0005610252,0.0001760805,0.005261947,0.001073597,0.0001656897,0.0005964122,0.0002586407,0.03354993,0.09839428,0.03000199,0.829333,0.0006273643],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"software","genre_gemma":"software","genre_scores_codex":[0.002458513,0.0002965585,0.2534351,0.0005888614,0.00036797,0.0004974583,0.04499031,0.6928097,0.00455554],"genre_scores_gemma":[0.02048361,0.0005580814,0.3250799,0.001770215,0.0001482403,0.003259474,0.1455885,0.4930097,0.01010239],"genre_candidate":"software","genre_consensus":"software","teacher_disagreement_score":0.9944023,"threshold_uncertainty_score":0.1726879,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09799458101637326,"score_gpt":0.4068712497562927,"score_spread":0.3088766687399195,"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."}}