Machine learning designs new GCGR/GLP-1R dual agonists with enhanced biological potency

• Published in: Nature chemistry Vol. 16; no. 9; pp. 1436 - 1444
• Main Authors: Puszkarska, Anna M., Taddese, Bruck, Revell, Jefferson, Davies, Graeme, Field, Joss, Hornigold, David C., Buchanan, Andrew, Vaughan, Tristan J., Colwell, Lucy J.
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 01-09-2024; Nature Publishing Group
• Subjects: 631/114/1305; 631/154/309/606; 631/154/51/2313; Agonists; Amino Acid Sequence; Analytical Chemistry; Biochemistry; Biological activity; Chemistry; Chemistry and Materials Science; Chemistry/Food Science; Cycle protein; Design; Design optimization; Diabetes mellitus (non-insulin dependent); Drug Design; Experimental data; G protein-coupled receptors; Glucagon; Glucagon-like peptide 1; Glucagon-Like Peptide-1 Receptor - agonists; Glucagon-Like Peptide-1 Receptor - metabolism; Humans; Hypoglycemic Agents - chemistry; Hypoglycemic Agents - pharmacology; Inorganic Chemistry; Learning algorithms; Ligands; Machine Learning; Neural networks; Organic Chemistry; Peptides; Peptides - chemistry; Peptides - pharmacology; Physical Chemistry; Protein engineering; Receptors; Receptors, Glucagon - agonists; Receptors, Glucagon - metabolism
• ISSN: 1755-4330; 1755-4349; 1755-4349
• DOI: 10.1038/s41557-024-01532-x

Several peptide dual agonists of the human glucagon receptor (GCGR) and the glucagon-like peptide-1 receptor (GLP-1R) are in development for the treatment of type 2 diabetes, obesity and their associated complications. Candidates must have high potency at both receptors, but it is unclear whether the limited experimental data available can be used to train models that accurately predict the activity at both receptors of new peptide variants. Here we use peptide sequence data labelled with in vitro potency at human GCGR and GLP-1R to train several models, including a deep multi-task neural-network model using multiple loss optimization. Model-guided sequence optimization was used to design three groups of peptide variants, with distinct ranges of predicted dual activity. We found that three of the model-designed sequences are potent dual agonists with superior biological activity. With our designs we were able to achieve up to sevenfold potency improvement at both receptors simultaneously compared to the best dual-agonist in the training set.
Engineering new ligands that specifically target multiple G protein-coupled receptors with desired activity profiles requires time-consuming and expensive cycles of design-make-test-analyse work. Now it has been shown that limited experimental data can be used to train sophisticated machine learning models to accurately predict the activity of previously uncharacterized peptide ligand variants.