{"id":"W1998262268","doi":"10.1016/j.tig.2007.07.008","title":"Reconsidering the significance of genomic word frequencies","year":2007,"lang":"en","type":"article","venue":"Trends in Genetics","topic":"RNA and protein synthesis mechanisms","field":"Biochemistry, Genetics and Molecular Biology","cited_by":42,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université de Montréal","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Biology; Genome; Computational biology; Copying; Genetics; Log-normal distribution; Evolutionary biology; Mathematics; Gene; Statistics","routes":{"ca_aff":true,"ca_fund":true,"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":[0.0004394706,0.00009665741,0.0001036117,0.00006248241,0.00003146729,0.000006797516,0.0002163102,0.00009058353,0.00004960022],"category_scores_gemma":[0.00002672899,0.00007830007,0.00005306177,0.0001060335,0.00008610838,9.317508e-7,0.0000539721,0.00005643989,0.000003248712],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000009822485,"about_ca_system_score_gemma":0.00002438759,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001091666,"about_ca_topic_score_gemma":0.0002904556,"domain_scores_codex":[0.9992582,0.00004026961,0.0002437429,0.0001748916,0.00008105693,0.0002019055],"domain_scores_gemma":[0.9994895,0.00003127082,0.00008390898,0.0003433207,0.00002335429,0.00002858869],"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":[0.00002197211,0.00001154921,0.001892466,0.000004532166,0.00001061269,0.000001857631,0.0001247679,0.0001793443,0.8961484,0.00006328774,0.00007390261,0.1014672],"study_design_scores_gemma":[0.0001237734,0.00007341076,0.01425973,0.00000675608,0.000006275285,0.000005624481,0.0001564945,0.0000158183,0.9783245,0.0005548079,0.006368284,0.000104491],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9929275,0.002277856,0.001674104,0.00007740988,0.0001210903,0.00007970168,0.00000477463,0.000004414882,0.00283317],"genre_scores_gemma":[0.9932923,0.0002369872,0.005676796,0.00006857868,0.00009859868,0.000009516511,0.00000476479,0.00001461832,0.0005978394],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1013627,"threshold_uncertainty_score":0.3192984,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06067486236173749,"score_gpt":0.269580137264331,"score_spread":0.2089052749025935,"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."}}