{"id":"W2153767201","doi":"10.1093/nar/gkl404","title":"Sequence biases in large scale gene expression profiling data","year":2006,"lang":"en","type":"article","venue":"Nucleic Acids Research","topic":"Gene expression and cancer classification","field":"Biochemistry, Genetics and Molecular Biology","cited_by":56,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canada's Michael Smith Genome Sciences Centre","funders":"","keywords":"Biology; Gene expression profiling; Gene; Serial analysis of gene expression; RefSeq; Genetics; Gene expression; Computational biology; DNA microarray; Sampling bias; Genome; Statistics; Sample size determination; Mathematics","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":[0.001101669,0.0001039531,0.000098489,0.0001446766,0.0001723458,0.00005168399,0.0007129157,0.0001590832,0.0000830179],"category_scores_gemma":[0.0002145431,0.00009462971,0.00002525045,0.0003518389,0.0001105647,0.00001504678,0.0006736238,0.0002289489,0.00004747153],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004095376,"about_ca_system_score_gemma":0.0001693323,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005925312,"about_ca_topic_score_gemma":0.0000913847,"domain_scores_codex":[0.9980073,0.0002203081,0.0002139195,0.000641137,0.0004267746,0.0004905579],"domain_scores_gemma":[0.9986163,0.00002228803,0.0000398035,0.00111163,0.0001305818,0.00007939262],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00005558437,0.0001070002,0.01880332,0.00001182785,0.000001599623,0.000004067754,0.00001898906,0.00004428832,0.9680968,0.00002781405,0.01178916,0.001039561],"study_design_scores_gemma":[0.0004589415,0.00005581495,0.008751084,0.00004716666,0.000001062189,0.00000396216,0.0002419302,0.001590126,0.9534475,0.00009774855,0.03518011,0.000124591],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9951337,0.0007434243,0.0007830089,0.0002121369,0.00006463123,0.000251799,0.00008769766,0.00001702929,0.002706557],"genre_scores_gemma":[0.9935803,0.0002139466,0.003387702,0.00004783258,0.0002969991,0.00005412786,0.001043469,0.00002380369,0.001351796],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02339095,"threshold_uncertainty_score":0.3858888,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1263237161734327,"score_gpt":0.4009265080186994,"score_spread":0.2746027918452668,"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."}}