Diffusion Probabilistic Modeling of Protein Backbones in 3D for the Motif-Scaffolding problem
Name
Yim-jyim-SM-EECS-2023-thesis.pdf
Description
Thesis PDF
Size
5.23 MB
Format
Adobe PDF
Checksum (MD5)
11fffb076451ae243e4a5af8ebddaa1a
Author(s)
Yim, Jason
Advisor(s)
Jaakkola, Tommi S.
Barzilay, Regina
Date Issued
February 2023
Publisher
Massachusetts Institute of Technology
Abstract
Construction of a scaffold structure that supports a desired motif, conferring protein function, shows promise for the design of vaccines and enzymes. But a general solution to this motif-scaffolding problem remains open. Current machine-learning techniques for scaffold design are either limited to unrealistically small scaffolds (up to length 20) or struggle to produce multiple diverse scaffolds. We propose to learn a distribution over diverse and longer protein backbone structures via an E(3)-equivariant graph neural network. We develop SMCDiff to efficiently sample scaffolds from this distribution conditioned on a given motif; our algorithm is the first to theoretically guarantee conditional samples from a diffusion model in the large-compute limit. We evaluate our designed backbones by how well they align with AlphaFold2-predicted structures. We show that our method can (1) sample scaffolds up to 80 residues and (2) achieve structurally diverse scaffolds for a fixed motif.
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
In Copyright - Educational Use Permitted
Copyright MIT
Persistent DSpace Link
