{"id":"W2745961286","doi":"10.1186/s13321-017-0234-y","title":"Comparative analysis of chemical similarity methods for modular natural products with a hypothetical structure enumeration algorithm","year":2017,"lang":"en","type":"article","venue":"Journal of Cheminformatics","topic":"Microbial Natural Products and Biosynthesis","field":"Medicine","cited_by":63,"is_retracted":false,"has_abstract":true,"ca_institutions":"MacEwan University; McMaster University","funders":"Canadian Institutes of Health Research; Canada Research Chairs","keywords":"Cheminformatics; Chemical space; Chemical similarity; Computer science; Natural product; Similarity (geometry); Enumeration; Modular design; Data mining; Drug discovery; Biochemical engineering; Algorithm; Structural similarity; Artificial intelligence; Bioinformatics; Chemistry; Mathematics; Biology; Engineering","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003909474,0.0007630051,0.0009047696,0.002636303,0.0006262022,0.001090964,0.001561961,0.001262862,0.002793585],"category_scores_gemma":[0.01491068,0.00022129,0.0009424965,0.002043373,0.0005622877,0.001813928,0.001177131,0.000736483,0.0004528137],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001317235,"about_ca_system_score_gemma":0.002043518,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002093469,"about_ca_topic_score_gemma":0.002692166,"domain_scores_codex":[0.9983513,0.0006771374,0.0001465334,0.0002497101,0.0004520559,0.000123326],"domain_scores_gemma":[0.9870783,0.01003813,0.0004999448,0.0008317114,0.00132275,0.0002291098],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001124177,0.0006261516,0.01007394,0.0004199287,0.0001946894,0.0001089909,0.0001270626,0.5921496,0.00677487,0.02546011,0.00324451,0.3596961],"study_design_scores_gemma":[0.00004006838,0.0001779174,0.0007300164,0.00001196669,0.00001689976,0.00004558343,0.00002933641,0.98965,0.002358761,0.006052182,0.0008776768,0.000009530044],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3869725,0.002365873,0.5973652,0.0004242213,0.0001089557,0.0003427359,0.0009530449,0.002780992,0.008686428],"genre_scores_gemma":[0.3974917,0.0004244516,0.5978351,0.0001702937,0.00004644156,0.0002559274,0.002125263,0.0002513584,0.001399464],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003909474,"threshold_uncertainty_score":0.02067548,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0263425672799847,"score_gpt":0.3490112789775224,"score_spread":0.3226687116975377,"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."}}