{"id":"W4414799063","doi":"10.1101/2025.10.02.25337130","title":"Formal Statistical Replication Analysis in Lung Cancer Genome-Wide Association Studies","year":2025,"lang":"en","type":"preprint","venue":"medRxiv","topic":"Bioinformatics and Genomic Networks","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Sinai Health System; Lunenfeld-Tanenbaum Research Institute; Public Health Ontario; University of Toronto","funders":"National Institutes of Health","keywords":"Replication (statistics); Lung cancer; False discovery rate; Spurious relationship; Single-nucleotide polymorphism; Genome-wide association study; SNP; Multiple comparisons problem; Genetic association","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":["metaresearch"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.3578926,0.002511252,0.003714018,0.004258516,0.00255685,0.004793233,0.007188171,0.003825201,0.006029175],"category_scores_gemma":[0.6528099,0.002153315,0.006638418,0.004563483,0.0101479,0.0059883,0.007000387,0.006196424,0.0009620967],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003402035,"about_ca_system_score_gemma":0.009499665,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004462157,"about_ca_topic_score_gemma":0.003123585,"domain_scores_codex":[0.6393386,0.3166839,0.01079565,0.01608852,0.0151235,0.001969896],"domain_scores_gemma":[0.2304934,0.6567089,0.02483421,0.07115763,0.01536388,0.001441995],"domain_codex":null,"domain_gemma":"reproducibility","domain_candidate":"reproducibility","domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"observational","study_design_scores_codex":[0.001904738,0.0003605205,0.03451971,0.003644296,0.005173694,0.00294841,0.002926468,0.1553405,0.003100624,0.6737219,0.01077452,0.1055845],"study_design_scores_gemma":[0.000862281,0.000381609,0.001963516,0.0002930261,0.0005913922,0.0004510014,0.0001994817,0.3194756,0.001520528,0.6694872,0.004658676,0.0001156922],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.007402572,0.0007993928,0.9872573,0.001547675,0.0003315055,0.0007752677,0.0003660744,0.000658077,0.0008620833],"genre_scores_gemma":[0.3564076,0.0005166219,0.6337849,0.001480796,0.000481803,0.005537406,0.0007945346,0.000273397,0.0007229175],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6421074,"threshold_uncertainty_score":0.7918327,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01282469955548649,"score_gpt":0.3094883562238908,"score_spread":0.2966636566684043,"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."}}