{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003020527,0.0004087998,0.0006588415,0.001919795,0.0009240025,0.004653663,0.001753906,0.001444765,0.009710925],"category_scores_gemma":[0.04051635,0.0003723485,0.0002892019,0.002002873,0.003656543,0.006070656,0.001491025,0.002282836,0.001336431],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007935465,"about_ca_system_score_gemma":0.0008320576,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002360204,"about_ca_topic_score_gemma":0.002753997,"domain_scores_codex":[0.998243,0.0006700009,0.00009351777,0.0004956275,0.0003722725,0.0001256802],"domain_scores_gemma":[0.9820439,0.01311166,0.0007923916,0.001521122,0.002016898,0.0005140403],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0004825181,0.00008322274,0.04094655,0.0003800228,0.0001862133,0.0005836771,0.003487231,0.003689655,0.02553226,0.7246714,0.00961991,0.1903375],"study_design_scores_gemma":[0.00005922179,0.00008617056,0.03437485,0.0001204093,0.000123456,0.000742197,0.00222495,0.01131753,0.006461487,0.9232349,0.0211414,0.0001133596],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6956183,0.009734228,0.1249982,0.05682411,0.003327604,0.00005076599,0.001782314,0.0007196267,0.1069449],"genre_scores_gemma":[0.97909,0.001485808,0.01313138,0.001680618,0.001029638,0.00002728918,0.0002936602,0.0002619483,0.002999608],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.009710925,"threshold_uncertainty_score":0.03248632,"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."}}