Design of peptide binders towards human cancer surface receptors
Abstract
Structure prediction technology has revolutionised the field of protein design, but key questions such as how to design new functions remain. Many proteins exercise their functions through interactions with other proteins and a significant challenge that remains is how well these interactions can be designed. Most efforts have focused on larger, more stable proteins, but shorter peptides are both cheaper to manufacture and are less sterically hindered. Previously, we have developed a design pipeline based on MC search through AlphaFold2 to design peptide binders called EvoBind (https://www.biorxiv.org/content/10.1101/2022.07.23.501214v1).
EvoBind is used by researchers all over the world and the loss function used has been independently validated through laboratory affinity analysis (https://www.nature.com/articles/s42256-023-00691-9). Now, we seek to develop EvoBind further by designing entire sequences at once using gradient descent through a trained structure prediction network (https://www.biorxiv.org/content/10.1101/2023.07.04.547638v1). We will then apply EvoBind to design peptide binders towards 400 human cancer cell surface receptors and evaluate the designs in the lab using surface plasmon resonance.
The aim is to create a wide range of peptides that can target a variety of different cancer types for both diagnostic and targeted therapy. This will be a first and help to provide a more personalised cancer treatment that can be adjusted as the cancer evolves following changes in the expression of different cell surface receptors. We foresee a growing market for peptide drugs in cancer treatment and diagnostics powered by binder design and new AI tools.
