{"id":"W2318488401","doi":"10.1093/bioinformatics/btw155","title":"MOLGENIS/connect: a system for semi-automatic integration of heterogeneous phenotype data with applications in biobanks","year":2016,"lang":"en","type":"article","venue":"Bioinformatics","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":18,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; McGill University Health Centre","funders":"Terveyden ja hyvinvoinnin laitos; Nederlandse Organisatie voor Wetenschappelijk Onderzoek; European Commission","keywords":"Computer science; Biobank; Documentation; Source code; Data integration; Data mining; Terminology; Open source; Information retrieval; Ontology; Categorical variable; Task (project management); Data source; Matching (statistics); Software; Programming language; Machine learning; 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":[],"consensus_categories":[],"category_scores_codex":[0.006365995,0.002132892,0.001245041,0.005397011,0.0009174209,0.002911986,0.002851105,0.001258998,0.02529471],"category_scores_gemma":[0.0186404,0.001480144,0.001695424,0.003537375,0.0007676438,0.003071574,0.005656826,0.001284304,0.0100332],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001352452,"about_ca_system_score_gemma":0.002277074,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002413324,"about_ca_topic_score_gemma":0.002734299,"domain_scores_codex":[0.9968316,0.0007436209,0.0004223495,0.001133636,0.0007426118,0.0001262927],"domain_scores_gemma":[0.991766,0.004494541,0.0009324845,0.001666647,0.0007109348,0.0004294355],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.004415947,0.0004507673,0.02808722,0.003324567,0.001178539,0.002266165,0.001857902,0.01200799,0.03189338,0.01401384,0.5380815,0.3624222],"study_design_scores_gemma":[0.001591229,0.0004652497,0.03318496,0.00110158,0.0005203587,0.003709823,0.0009076775,0.2727322,0.1002689,0.05247515,0.5325004,0.0005424732],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"software","genre_gemma":"software","genre_scores_codex":[0.02013474,0.0006638666,0.3587037,0.001019255,0.0002533598,0.001130476,0.04997556,0.5613034,0.006815791],"genre_scores_gemma":[0.1110262,0.0006041285,0.688768,0.001902465,0.0002252928,0.00352455,0.1435439,0.04422567,0.00617973],"genre_candidate":"software","genre_consensus":"software","teacher_disagreement_score":0.02529471,"threshold_uncertainty_score":0.08461928,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02425168455166588,"score_gpt":0.2718240070330186,"score_spread":0.2475723224813527,"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."}}