{"id":"W2904022159","doi":"10.1101/500686","title":"Text-mining clinically relevant cancer biomarkers for curation into the CIViC database","year":2018,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University; Canada's Michael Smith Genome Sciences Centre; University of British Columbia","funders":"National Cancer Institute; National Human Genome Research Institute; National Institutes of Health","keywords":"Data curation; Computer science; Precision medicine; Construct (python library); Cancer; Resource (disambiguation); MEDLINE; Information retrieval; Data science; Medicine; Pathology; Biology; Internal medicine","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":[0.006856936,0.002189955,0.002515572,0.02786106,0.001647747,0.003686054,0.002992731,0.003000299,0.01365278],"category_scores_gemma":[0.0328677,0.0009960444,0.001915916,0.0158127,0.0006748128,0.003276172,0.003285807,0.002243345,0.0090236],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00251478,"about_ca_system_score_gemma":0.008056015,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005464245,"about_ca_topic_score_gemma":0.01082521,"domain_scores_codex":[0.9958912,0.0007236116,0.001368577,0.0009986868,0.0008573916,0.0001604851],"domain_scores_gemma":[0.9728276,0.01434396,0.004131169,0.002425086,0.005336606,0.0009354783],"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":[0.001258484,0.0006173234,0.0175493,0.04995839,0.001048923,0.006217251,0.002544673,0.005400245,0.03458298,0.01237898,0.6263517,0.2420917],"study_design_scores_gemma":[0.0004944646,0.0003864313,0.02368653,0.006927995,0.001678315,0.002615572,0.001568172,0.01937525,0.03055212,0.01315306,0.8992673,0.0002948828],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"methods","genre_scores_codex":[0.01504089,0.007543653,0.04089352,0.002754051,0.0004721525,0.002178254,0.9037026,0.01814788,0.009267001],"genre_scores_gemma":[0.02237536,0.003025987,0.1541324,0.001151975,0.000196715,0.002088374,0.8144001,0.000958549,0.001670456],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.02786106,"threshold_uncertainty_score":0.04567307,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02491469509453119,"score_gpt":0.2972968484878694,"score_spread":0.2723821533933382,"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."}}