Sequence-based drug design using transformers
Zhang, S.; Huo, D.; Horne, R. I.; Qi, Y.; Ojeda, S. P.; Yan, A.; Vendruscolo, M.
Show abstract
Protein-ligand interactions play central roles in biological processes and are of key importance in drug design. Deep learning-based approaches are emerging as cost-effective alternatives to high-throughput experimental methods for the screening of large libraries of ligands. Here, to predict the binding affinity between proteins and small molecules, we introduce Ligand-Transformer, a deep learning framework based on the AlphaFold2 transformer architecture. We applied Ligand-Transformer to screen inhibitors targeting the mutant EGFRLTC kinase, identifying compounds with low nanomolar potency. We then used this approach to predict the conformational population shifts induced by ABL kinase inhibitors. To show the applicability of Ligand-Transformer to disordered proteins, we explored the binding of small molecules to the Alzheimers A{beta} peptide, identifying compounds that delayed its aggregation. Overall, Ligand-Transformer illustrates the potential of transformers in accurately predicting the interactions of small molecules with both ordered and disordered proteins, thus uncovering molecular mechanisms and facilitating the initial steps in drug discovery.
Matching journals
The top 4 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.