{"id":"W4411721546","doi":"10.1101/2025.06.24.661356","title":"Better data for better predictions: data curation improves deep learning for sgRNA/Cas9 prediction","year":2025,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Advanced biosensing and bioanalysis techniques","field":"Biochemistry, Genetics and Molecular Biology","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Canadian Institutes of Health Research","keywords":"Computer science; Data curation; Artificial intelligence; Deep learning; CRISPR; Machine learning; Data science; Biology; Genetics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.005888779,0.001741948,0.0009256287,0.002678628,0.0008208972,0.002554562,0.002540542,0.001196955,0.004810239],"category_scores_gemma":[0.02729833,0.0007313291,0.001268934,0.002192797,0.000679771,0.003217652,0.002852038,0.00387873,0.00331832],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001141197,"about_ca_system_score_gemma":0.002863942,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01102694,"about_ca_topic_score_gemma":0.01954535,"domain_scores_codex":[0.9967366,0.0008868265,0.0002682597,0.000950499,0.0009983798,0.0001593938],"domain_scores_gemma":[0.9843172,0.006632315,0.0006389683,0.004751966,0.003223638,0.0004359726],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001296305,0.0009165262,0.03070716,0.001825338,0.001024107,0.0006117165,0.0003786852,0.1889133,0.03812458,0.01150059,0.2755433,0.4491583],"study_design_scores_gemma":[0.000137882,0.00009450992,0.003331539,0.0001942181,0.00008403658,0.00009150105,0.00007574975,0.8952306,0.03345731,0.01738146,0.04983319,0.00008807176],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0880295,0.002445138,0.7062464,0.006163792,0.001128658,0.0004191238,0.07226168,0.1173308,0.005974915],"genre_scores_gemma":[0.2203578,0.0007567805,0.6329259,0.001360085,0.0001859261,0.0004747097,0.1346012,0.005735638,0.003601877],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01102694,"threshold_uncertainty_score":0.03114319,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02202305941005196,"score_gpt":0.2773886235241835,"score_spread":0.2553655641141316,"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."}}