{"id":"W6921191962","doi":"10.6084/m9.figshare.25067253.v1","title":"Additional file 1 of Anchor Clustering for million-scale immune repertoire sequencing data","year":2024,"lang":"en","type":"dataset","venue":"Figshare","topic":"Mathematics Education and Programs","field":"Mathematics","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph","funders":"","keywords":"Cluster analysis; Repertoire; DNA sequencing; Key (lock); Data file","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.002230388,0.001810664,0.001529343,0.003375284,0.001428508,0.002603184,0.00270391,0.001805575,0.5525054],"category_scores_gemma":[0.01591419,0.0008126856,0.001351423,0.005419901,0.0005608436,0.001677591,0.001709437,0.002160804,0.2027621],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001214519,"about_ca_system_score_gemma":0.002570714,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007955045,"about_ca_topic_score_gemma":0.02009282,"domain_scores_codex":[0.9988737,0.0001773549,0.0001151893,0.000450374,0.000201219,0.000182238],"domain_scores_gemma":[0.9918289,0.005374098,0.0004258877,0.000956137,0.0009906286,0.0004244604],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000148405,0.0000580804,0.002127774,0.001457115,0.00007090491,0.0000352039,0.00005326596,0.0005431885,0.0003047273,0.000599041,0.9908665,0.003735691],"study_design_scores_gemma":[0.00135186,0.0001149552,0.01414695,0.0009061772,0.00020227,0.0002836962,0.0001985027,0.00147861,0.001182429,0.006767144,0.9732606,0.0001067173],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.00008884278,0.0000186764,0.0001610634,0.00002086806,0.00001157325,0.00001888814,0.9991399,0.0002762757,0.0002639441],"genre_scores_gemma":[0.001123042,0.00003521728,0.001343164,0.00009897012,0.00001339153,0.0002910692,0.9955555,0.0004337551,0.001105921],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.5525054,"threshold_uncertainty_score":0.6382967,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1923862650353638,"score_gpt":0.3682997128319365,"score_spread":0.1759134477965727,"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."}}