{"id":"W4416947530","doi":"10.1038/s41597-025-06194-z","title":"A Unified Cellular Toxic Potency Dataset of PM2.5 across Chemicals, Emission Sources, and Regions in China","year":2025,"lang":"en","type":"article","venue":"Scientific Data","topic":"Air Quality and Health Impacts","field":"Environmental Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Environment and Climate Change Canada","funders":"Fudan University; National Natural Science Foundation of China","keywords":"Human health; Risk assessment; Health risk; Resource (disambiguation); China; Particulates; Health risk assessment; Bioassay; Air quality index","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":[],"domain":null,"study_design":"not_applicable","genre":"dataset","about_ca_system":false,"about_ca_topic":false,"confidence":"low","status":"direct model label, unvalidated"}],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0008543219,0.001202012,0.001201656,0.003366074,0.0007484761,0.0008652473,0.001328651,0.0008187851,0.001388718],"category_scores_gemma":[0.001194302,0.000283092,0.001160435,0.005641242,0.0004374278,0.0005209951,0.001413304,0.0003822595,0.0006281946],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002574464,"about_ca_system_score_gemma":0.003785195,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.1096377,"about_ca_topic_score_gemma":0.08337714,"domain_scores_codex":[0.9990606,0.0001121356,0.0001181306,0.0003037163,0.0002667847,0.0001387066],"domain_scores_gemma":[0.9991364,0.0001052217,0.00009257417,0.0002056572,0.0003650017,0.0000951133],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.001019325,0.0005575097,0.6612575,0.004081485,0.001267311,0.001667745,0.0005974837,0.0719422,0.02773775,0.003566655,0.1113564,0.1149486],"study_design_scores_gemma":[0.0001720737,0.0002428143,0.8368837,0.0001841735,0.0005406597,0.0004497748,0.0004601649,0.03624973,0.01036618,0.001960061,0.1123283,0.0001623624],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.364941,0.002160229,0.004687367,0.0003585353,0.00006399106,0.0002843791,0.6214863,0.0007528354,0.005265471],"genre_scores_gemma":[0.3135745,0.0009032812,0.002676326,0.0001317654,0.00002985233,0.0003672477,0.6811095,0.00003480219,0.001172715],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.1096377,"threshold_uncertainty_score":0.217999,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05077983788302821,"score_gpt":0.3439604614852171,"score_spread":0.2931806236021889,"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."}}