{"id":"W4389174422","doi":"10.1038/s41588-023-01524-6","title":"Benchmarking of deep neural networks for predicting personal gene expression from DNA sequence highlights shortcomings","year":2023,"lang":"en","type":"article","venue":"Nature Genetics","topic":"Genomics and Chromatin Dynamics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":114,"is_retracted":false,"has_abstract":false,"ca_institutions":"Canadian Institute for Advanced Research","funders":"National Institute on Aging","keywords":"Benchmarking; Biology; Computational biology; DNA sequencing; Gene; Genetics; Personal genomics; Genome; Genomics; Human genome","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001444342,0.000197695,0.0001892168,0.00004488305,0.0001268064,0.0000244233,0.0002934869,0.0005357765,0.000006059497],"category_scores_gemma":[0.00004399427,0.0001928119,0.0001375666,0.0001186048,0.00005153145,0.000002956923,0.0002018175,0.0002052806,4.552481e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001716616,"about_ca_system_score_gemma":0.00002918608,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000004038604,"about_ca_topic_score_gemma":0.00002463264,"domain_scores_codex":[0.9987448,0.00002599871,0.0002877015,0.0004345947,0.000177754,0.0003291884],"domain_scores_gemma":[0.9992862,0.00004738739,0.0001706821,0.000292386,0.0001233978,0.00007995794],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00005643662,0.00001724057,0.009257891,0.0000289337,0.0000520668,0.00000345483,0.0001763384,0.01354002,0.9725799,0.000008574819,0.0004238061,0.003855339],"study_design_scores_gemma":[0.0003923516,0.0001563238,0.006217779,0.00002314745,0.00004064973,0.000005905592,0.00006808139,0.6660666,0.3257343,0.0001195011,0.0009421043,0.0002332565],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9861739,0.001841625,0.01077034,0.00005686812,0.0006832795,0.0002308912,0.0001902143,0.00002341065,0.00002949637],"genre_scores_gemma":[0.9811576,0.0003530699,0.01573377,0.0001170164,0.00101589,0.00002076016,0.001529735,0.00004191998,0.00003022894],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6525266,"threshold_uncertainty_score":0.7862641,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.008019904422525752,"score_gpt":0.2400583485352147,"score_spread":0.232038444112689,"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."}}