{"id":"W7085190714","doi":"10.6084/m9.figshare.30286912","title":"Comprehensive Global Dataset of SARS-CoV-2 Lineages and Spike Mutations for Therapeutic and Vaccine Research (734 Cities, 2019 - 2025 Q1, n=9.4 Million Approx.)","year":2025,"lang":"en","type":"dataset","venue":"Figshare","topic":"Neonatal and Maternal Infections","field":"Medicine","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Spike (software development); Spike train; Mutation; Quarter (Canadian coin); Genomics; Feature (linguistics)","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.00008302617,0.0002624291,0.0005130403,0.0002671946,0.0001667299,0.00005457394,0.0001657083,0.0002611175,0.001674617],"category_scores_gemma":[0.0003059847,0.0002247279,0.00006974979,0.0003092932,0.00004788896,0.00008477495,0.0003471167,0.0003826626,0.00003310699],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005170011,"about_ca_system_score_gemma":0.0002204978,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000725363,"about_ca_topic_score_gemma":0.0002200674,"domain_scores_codex":[0.9986055,0.00006537137,0.000362176,0.0004333763,0.0002595422,0.0002740362],"domain_scores_gemma":[0.9982934,0.0004990951,0.0001420074,0.0004639018,0.0005237153,0.00007787507],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0001446594,0.00008946913,0.000007895398,0.006055645,0.0001818145,0.00002143255,0.000007622979,4.127286e-7,0.0001677211,0.00001026799,0.992301,0.001012039],"study_design_scores_gemma":[0.00114227,0.0003733594,0.0004352561,0.003319518,0.0002080548,0.00008153103,0.00003702784,0.00003356592,0.0007255797,0.0001658129,0.9933079,0.000170107],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.0001229638,0.007248381,0.000003609756,0.0002936795,0.0001051819,0.001329831,0.9908429,0.00001938058,0.00003404042],"genre_scores_gemma":[0.00003742267,0.0009255713,0.0000842912,0.0004676558,0.0001668752,0.0003183657,0.9976233,0.00001088088,0.0003656968],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.006780309,"threshold_uncertainty_score":0.999238,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.163628285429014,"score_gpt":0.4407021651989538,"score_spread":0.2770738797699397,"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."}}