{"id":"W1589570762","doi":"10.1023/a:1026156025612","title":"Coverage of protein sequence space by current structural genomics targets","year":2003,"lang":"en","type":"article","venue":"Journal of Structural and Functional Genomics","topic":"Enzyme Structure and Function","field":"Materials Science","cited_by":13,"is_retracted":false,"has_abstract":false,"ca_institutions":"Biotechnology Research Institute; McGill University","funders":"Canadian Institutes of Health Research","keywords":"Structural genomics; Protein Data Bank (RCSB PDB); Proteome; Computational biology; Protein superfamily; Homology (biology); Biology; Protein Data Bank; Homologous chromosome; Genome; Cluster analysis; Proteomics; Genetics; Protein structure; Computer science; Gene; Biochemistry; Artificial intelligence","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002798285,0.0002369982,0.0003730705,0.00009064058,0.0001843254,0.00006519064,0.0001390757,0.00009852081,0.0004040542],"category_scores_gemma":[0.00007842104,0.0001778497,0.000123301,0.0001075876,0.0001745189,0.0003888603,0.00003956146,0.0003090268,0.000003730679],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001617428,"about_ca_system_score_gemma":0.0001846696,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001141638,"about_ca_topic_score_gemma":0.000002079255,"domain_scores_codex":[0.9984174,0.00007625591,0.0006455854,0.0002330259,0.0003692324,0.0002584419],"domain_scores_gemma":[0.9986385,0.00004670967,0.0007425194,0.0001148392,0.000299673,0.0001577557],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0002902031,0.000007483204,0.0008202189,0.00005267447,0.00003122685,0.000004646365,0.0001455413,0.0009503902,0.9904155,0.004498009,0.0003713691,0.002412807],"study_design_scores_gemma":[0.003316471,0.0009800347,0.021645,0.00009221066,0.000176513,0.002737361,0.0003448401,0.000478942,0.8603922,0.06441714,0.04452027,0.000899029],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9918141,0.004028867,0.001744803,0.00006146766,0.001867547,0.0001508487,0.0001324847,0.000008440724,0.0001914607],"genre_scores_gemma":[0.9954467,0.000215034,0.003775519,0.00004693294,0.0003758161,0.000001479991,0.00001282882,0.00001642269,0.0001092001],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1300232,"threshold_uncertainty_score":0.7252498,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01503592279387621,"score_gpt":0.2281130281210093,"score_spread":0.213077105327133,"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."}}