{"id":"W4411397654","doi":"10.1101/2025.06.16.659836","title":"SARS-CoV-2 ORF8 sequence conservation and mutational analysis — insight into the influence of dataset size on identifying top mutations","year":2025,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"SARS-CoV-2 and COVID-19 Research","field":"Medicine","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Trinity College","funders":"","keywords":"Sequence (biology); Coronavirus disease 2019 (COVID-19); Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2); Mutation; 2019-20 coronavirus outbreak; Sequence analysis; Biology; Genetics; Virology; Medicine; Gene; Internal medicine","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":[],"consensus_categories":[],"category_scores_codex":[0.007981872,0.0007631275,0.001090118,0.002618846,0.000806183,0.001652171,0.0008857534,0.0007319702,0.0007051919],"category_scores_gemma":[0.02241676,0.0002411396,0.001244986,0.001512443,0.0005366821,0.001724304,0.0008745181,0.00115752,0.0004171237],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000428371,"about_ca_system_score_gemma":0.0008326684,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002398645,"about_ca_topic_score_gemma":0.004248737,"domain_scores_codex":[0.9957778,0.001451948,0.0005954393,0.001170788,0.0007364364,0.0002674749],"domain_scores_gemma":[0.985041,0.009897354,0.001136888,0.001928154,0.001385713,0.0006107948],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.003925509,0.001259344,0.6675816,0.002017829,0.00238938,0.001282918,0.0007833486,0.06837729,0.08922006,0.001604743,0.01724213,0.1443159],"study_design_scores_gemma":[0.0002508749,0.001620182,0.3651498,0.0003274047,0.0008205804,0.001660512,0.001321538,0.5775325,0.03437883,0.003957159,0.0128263,0.0001542412],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9776722,0.002483722,0.008915531,0.0005175198,0.0002045174,0.00009462249,0.008035873,0.0009848847,0.001091062],"genre_scores_gemma":[0.9546242,0.000406178,0.02279774,0.0001823932,0.00006353777,0.00009385539,0.02130643,0.0002123257,0.0003134749],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.007981872,"threshold_uncertainty_score":0.04221267,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04945484473830473,"score_gpt":0.3450903254181797,"score_spread":0.295635480679875,"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."}}