{"id":"W6902223625","doi":"10.6084/m9.figshare.28059550.v1","title":"Additional file 1 of CGRclust: Chaos Game Representation for twin contrastive clustering of unlabelled DNA sequences","year":2024,"lang":"en","type":"article","venue":"Figshare","topic":"Fractal and DNA sequence analysis","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Western University; University of Waterloo","funders":"","keywords":"Cluster analysis; Representation (politics); CHAOS (operating system); Pattern recognition (psychology); Feature (linguistics)","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":[0.0007350004,0.001636104,0.001302799,0.001564043,0.0007601177,0.001874879,0.002549578,0.001444799,0.698034],"category_scores_gemma":[0.009321098,0.0007625438,0.000901284,0.001918812,0.000360556,0.001252716,0.001177493,0.001167565,0.1656465],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001082487,"about_ca_system_score_gemma":0.001022276,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003685723,"about_ca_topic_score_gemma":0.007438118,"domain_scores_codex":[0.9996161,0.00005817944,0.00002878984,0.0001122437,0.0001218248,0.00006273012],"domain_scores_gemma":[0.9971245,0.002129792,0.0001035294,0.0002425302,0.0002771951,0.000122452],"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":[0.0003337431,0.0001051901,0.00137394,0.001887681,0.00006803565,0.0001296724,0.00009012915,0.005537754,0.001164265,0.003889845,0.9703461,0.0150737],"study_design_scores_gemma":[0.002875848,0.0003282942,0.008589214,0.001074649,0.0002241797,0.0006529432,0.0002734757,0.08397839,0.01026143,0.06337683,0.8280749,0.000289902],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.001242339,0.00007408604,0.008745014,0.0001626768,0.0001033972,0.0001021685,0.9678201,0.01793897,0.003811266],"genre_scores_gemma":[0.0330666,0.0002004111,0.04056145,0.0004569896,0.0001143918,0.001448804,0.8904341,0.02243637,0.01128084],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.698034,"threshold_uncertainty_score":0.4307178,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02372716648541075,"score_gpt":0.2767611067718642,"score_spread":0.2530339402864534,"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."}}