De Novo Design of High-Affinity Binders for Bioactive Helical Peptides

Explore cutting-edge techniques for designing high-affinity binders of bioactive helical peptides in this 37-minute conference talk by Susana Vazquez Torres from Valence Labs. Delve into advanced methods like parametric generation and deep learning-based approaches for protein design, with a focus on addressing the challenge of designing interactions between proteins and short peptides with helical propensity. Learn about the extension of RFdiffusion to enable binder design for flexible targets and the refinement of input structure models through successive noising and denoising. Discover how these techniques can generate picomolar-affinity binders to helical peptide targets, either by refining existing designs or creating them de novo from random noise distributions. Understand the practical applications of these designs in enriching and detecting hormones like parathyroid hormone and glucagon through mass spectrometry, as well as constructing bioluminescence-based protein biosensors. Gain insights into the broader implications of designing binders for conformationally variable targets and optimizing both natural and designed proteins through partial diffusion.

- Intro + Background
- Parametric Design
- RFJoint Inpainting + ProteinMPNN
- AF2 Hallucination
- RFDiffusion
- Partial Diffusion
- Peptide Binders as Diagnostic Tools
- Conclusion
- Q+A