{"id":"W4316495494","doi":"10.12688/wellcomeopenres.18681.1","title":"Pf7: an open dataset of Plasmodium falciparum genome variation in 20,000 worldwide samples","year":2023,"lang":"en","type":"preprint","venue":"Wellcome Open Research","topic":"Malaria Research and Control","field":"Medicine","cited_by":148,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary","funders":"Faculty of Tropical Medicine, Mahidol University; Addis Ababa University; National Institutes of Health; Universiteit Antwerpen; National Institute for Medical Research; Ministère de la Santé; Universidad Nacional de Colombia; Bundesministerium für Gesundheit; World Health Organization; University of Ghana; Bill and Melinda Gates Foundation; Mahidol University; University College London; National Center for Advancing Translational Sciences; National Human Genome Research Institute; Wellcome Trust; National Institute of Allergy and Infectious Diseases; Medical Research Council; London School of Hygiene and Tropical Medicine","keywords":"Plasmodium falciparum; Variation (astronomy); Biology; Genome; Computational biology; Structural variation; Genetics; Malaria; Evolutionary biology; Gene; Immunology","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[{"model":"gemma","categories":[],"domain":null,"study_design":"not_applicable","genre":"dataset","about_ca_system":false,"about_ca_topic":false,"confidence":"low","status":"direct model label, unvalidated"},{"model":"gpt","categories":["open_science"],"domain":null,"study_design":"not_applicable","genre":"dataset","about_ca_system":false,"about_ca_topic":false,"confidence":"medium","status":"direct model label, unvalidated"}],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","scholarly_communication","open_science","research_integrity","insufficient_payload"],"consensus_categories":["open_science"],"category_scores_codex":[0.01816784,0.000439904,0.001575739,0.001918708,0.00023526,0.001359556,0.008689972,0.0005284358,0.004493767],"category_scores_gemma":[0.001745661,0.0004068643,0.0001064208,0.001460413,0.0002963298,0.0005691056,0.02100332,0.003414029,0.0007237471],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005269145,"about_ca_system_score_gemma":0.002961498,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0900839,"about_ca_topic_score_gemma":0.01431276,"domain_scores_codex":[0.990455,0.002302214,0.001351933,0.001817248,0.002458382,0.001615211],"domain_scores_gemma":[0.9934971,0.001107666,0.0002981749,0.003586749,0.0006885637,0.0008217653],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"observational","study_design_scores_codex":[0.08741723,0.01809392,0.06750414,0.01517068,0.005315058,0.006949323,0.006867888,0.008528499,0.1450857,0.05927054,0.5502121,0.02958498],"study_design_scores_gemma":[0.01785034,0.003260144,0.6207926,0.0026003,0.0001736963,0.00002724736,0.001124299,0.01474427,0.001210906,0.02892185,0.307861,0.001433397],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.3634628,0.002403042,0.001929642,0.16551,0.002311046,0.1287945,0.1951501,0.0004509519,0.139988],"genre_scores_gemma":[0.8747709,0.001327725,0.005172191,0.0003642733,0.0007022879,0.005269598,0.0807943,0.0002836944,0.03131505],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5532884,"threshold_uncertainty_score":0.9998383,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2762833932571187,"score_gpt":0.4550705055030257,"score_spread":0.178787112245907,"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."}}