{"id":"W3157378145","doi":"10.48550/arxiv.2104.14661","title":"Random Embeddings and Linear Regression can Predict Protein Function","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Microplastics and Plastic Pollution","field":"Environmental Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Linear regression; Regression; Function (biology); Mathematics; Statistics; Econometrics; Statistical physics; Physics; Biology; Evolutionary biology","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":[],"consensus_categories":[],"category_scores_codex":[0.002201422,0.002236261,0.0009892164,0.001113869,0.0002844209,0.001081887,0.001052786,0.001733208,0.00222662],"category_scores_gemma":[0.008705045,0.0005250916,0.001030037,0.0008607124,0.0009609638,0.002797315,0.001328052,0.002757974,0.002760788],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007091438,"about_ca_system_score_gemma":0.0004802018,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001634145,"about_ca_topic_score_gemma":0.001978772,"domain_scores_codex":[0.9987258,0.0005155942,0.00005509888,0.0004544047,0.0001616519,0.00008738498],"domain_scores_gemma":[0.9963453,0.002116437,0.0003218073,0.0007215222,0.0003747283,0.0001202527],"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":[0.0008653832,0.0004593807,0.009149543,0.000599562,0.0004846769,0.0001697493,0.0001369807,0.5227966,0.0112263,0.01550029,0.02983004,0.4087815],"study_design_scores_gemma":[0.00002014637,0.00007877743,0.0008182398,0.00002282538,0.00001819299,0.00002758881,0.00001333916,0.9758705,0.001982755,0.01989648,0.001235218,0.00001591482],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2760531,0.009237741,0.6843503,0.003556994,0.0007314382,0.000121928,0.003490112,0.01421229,0.008246196],"genre_scores_gemma":[0.8806283,0.001880069,0.09777897,0.00083297,0.0003952796,0.0001502365,0.00953842,0.0008048334,0.007991015],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002236261,"threshold_uncertainty_score":0.0116424,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02281780920656047,"score_gpt":0.1562756846970882,"score_spread":0.1334578754905277,"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."}}