{"id":"W2147702777","doi":"10.1093/bioinformatics/btr229","title":"Sequence-based prediction of protein crystallization, purification and production propensity","year":2011,"lang":"en","type":"article","venue":"Bioinformatics","topic":"Enzyme Structure and Function","field":"Materials Science","cited_by":77,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"Natural Sciences and Engineering Research Council of Canada; Killam Trusts","keywords":"Crystallization; Protein crystallization; Computer science; Process (computing); Production (economics); Set (abstract data type); In silico; Selection (genetic algorithm); Sequence (biology); Quality (philosophy); Diffraction; Algorithm; Data mining; Biological system; Artificial intelligence; Chemistry; Biology; Physics; Genetics","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":[0.002578736,0.0008947665,0.000775877,0.00113371,0.0002724851,0.0008824441,0.0006442162,0.0007570505,0.0009047618],"category_scores_gemma":[0.00706076,0.0003071006,0.0005149004,0.0007918737,0.0005098993,0.001011679,0.0005692315,0.0008731122,0.0008041791],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006408002,"about_ca_system_score_gemma":0.001107865,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002724764,"about_ca_topic_score_gemma":0.002369912,"domain_scores_codex":[0.9991571,0.0001857288,0.00007886107,0.0002775367,0.0002317308,0.000069052],"domain_scores_gemma":[0.9949123,0.00296663,0.0008355272,0.0003090849,0.0007856495,0.0001908156],"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":[0.000656948,0.000336585,0.2495345,0.0007263597,0.0002202354,0.0003705374,0.0001235431,0.5815008,0.03225296,0.001568928,0.003538573,0.12917],"study_design_scores_gemma":[0.00001755386,0.0001642583,0.02418774,0.00003138983,0.00004006938,0.0002328644,0.00001923441,0.9504293,0.0224768,0.001515937,0.0008613038,0.00002360257],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5867427,0.002062556,0.4034823,0.0004669531,0.00003450617,0.0001538212,0.002437853,0.002957978,0.001661423],"genre_scores_gemma":[0.9224925,0.0006025177,0.07263508,0.0000926382,0.00003749352,0.0001038913,0.003211637,0.00009280754,0.0007314226],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.002724764,"threshold_uncertainty_score":0.01363784,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0539457064322674,"score_gpt":0.2132408168633426,"score_spread":0.1592951104310752,"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."}}