{"id":"W2573117738","doi":"10.1093/database/baw147","title":"The BioC-BioGRID corpus: full text articles annotated for curation of protein–protein and genetic interactions","year":2016,"lang":"en","type":"article","venue":"Database","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":29,"is_retracted":false,"has_abstract":true,"ca_institutions":"Mount Sinai Hospital; Lunenfeld-Tanenbaum Research Institute; Université de Montréal; Institute for Research in Immunology and Cancer","funders":"Biotechnology and Biological Sciences Research Council; National Institutes of Health","keywords":"Data curation; Annotation; Crowdsourcing; Computer science; Task (project management); Information retrieval; Gene Annotation; Pipeline (software); World Wide Web; Data science; Natural language processing; Computational biology; Artificial intelligence; Biology; Gene; Genome; Genetics; Engineering","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":[0.003098526,0.001641422,0.001361466,0.02126529,0.002784988,0.003135238,0.001989264,0.002031925,0.02024647],"category_scores_gemma":[0.01620208,0.0006109745,0.0007237725,0.02609695,0.001689674,0.002063732,0.003615979,0.001786069,0.01452683],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00248174,"about_ca_system_score_gemma":0.005512308,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0101502,"about_ca_topic_score_gemma":0.0282581,"domain_scores_codex":[0.9962913,0.0007727941,0.0008448606,0.0007200956,0.001172252,0.0001986981],"domain_scores_gemma":[0.9774449,0.01472818,0.001787689,0.001499842,0.003728505,0.0008109412],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0004863683,0.0001331247,0.002649002,0.02083196,0.0001525652,0.002653658,0.002293893,0.0007540624,0.01400317,0.003926239,0.8940946,0.05802127],"study_design_scores_gemma":[0.0001406641,0.00003311234,0.01221361,0.001660362,0.00013472,0.0009985316,0.001091932,0.0007339618,0.004950676,0.001208884,0.9767725,0.00006095893],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"methods","genre_scores_codex":[0.02363438,0.01364916,0.007220298,0.002476092,0.0009461161,0.001017946,0.9286564,0.003011658,0.01938801],"genre_scores_gemma":[0.02132499,0.004333302,0.01519106,0.0005692546,0.0002523643,0.001730275,0.9503819,0.0009158172,0.005300917],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.02126529,"threshold_uncertainty_score":0.0677312,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0194035114125649,"score_gpt":0.2759297630851905,"score_spread":0.2565262516726256,"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."}}