{"id":"W4412146709","doi":"10.1039/d5dd00107b","title":"Chemical language models can generate biomolecules atom-by-atom","year":2025,"lang":"en","type":"article","venue":"Digital Discovery","topic":"Machine Learning in Materials Science","field":"Materials Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canadian Institute for Advanced Research; Vector Institute; University of Toronto","funders":"","keywords":"Biomolecule; Atom (system on chip); Chemistry; Computer science; Nanotechnology; Materials science; Parallel computing","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":"codex-gemma-dda1882f352a","candidate_categories":["scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.000187669,0.000236001,0.0002555488,0.00007987807,0.0001066707,0.001647,0.0006397443,0.00007826918,0.00009626472],"category_scores_gemma":[0.00008243068,0.0002022518,0.00007968589,0.000279554,0.0002434106,0.001250277,0.0003845681,0.0000966564,0.0001627303],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007262725,"about_ca_system_score_gemma":0.0001195328,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001828623,"about_ca_topic_score_gemma":0.000006008365,"domain_scores_codex":[0.9982734,0.00005074157,0.000306098,0.0005731467,0.0003379069,0.0004587256],"domain_scores_gemma":[0.9992754,0.00006195813,0.00009105202,0.0004394568,0.00003634676,0.00009571658],"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":[0.00002095633,0.00005685234,0.0002265329,0.0000306366,0.000005728951,0.0000159272,0.0001376866,0.0006067444,0.9836239,0.01222087,0.002561729,0.0004924985],"study_design_scores_gemma":[0.0002920455,0.00001785404,0.00006892635,0.00006327809,0.00001079021,0.00001047537,0.0001178999,0.00395754,0.9879047,0.004438162,0.002741277,0.0003770565],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9615196,0.0002023714,0.02091416,0.000336263,0.0004569856,0.0001270817,0.0005894752,0.0001830786,0.015671],"genre_scores_gemma":[0.9938123,0.000004840309,0.001026792,0.0006093568,0.00006075874,0.00002104883,0.0001314855,0.00002152592,0.004311846],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.03229275,"threshold_uncertainty_score":0.9993894,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.005876280322459212,"score_gpt":0.2430713952140933,"score_spread":0.2371951148916341,"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."}}