Deep learning for protein design
Abstract/Contents
- Abstract
- Advancements in computational modeling methods have led to remarkable outcomes in protein design, including the development of new therapeutics, enzymes, and biosensors. However, the protein design problem remains a major engineering challenge, as the current design process relies heavily on heuristics, requiring subjective expertise to negotiate pitfalls that result from optimizing imperfect scoring functions. The ability to quickly and accurately generate protein structures and sequences would constitute a major advancement in the field of protein design. Improved methods for function approximation, in particular deep learning, might allow us to move beyond the limitations of current methods. My thesis work explores the development of deep learning systems to learn and sample from the distributions of protein structures and sequences. I first describe my work on training deep generative models to generate protein backbones and will further describe applications of this method to loop modeling. Next, I describe efforts to move toward practicable protein backbone generation including generating backbone coordinates in a fully differentiable manner, as well as generating domain-specific backbones or backbones conditioned on properties of interest. Next, I describe an entirely learned method for protein sequence design on fixed protein backbones, highlighting some interesting experimental results validating the method. Finally, I conclude with a perspective on future directions in this field.
Description
| Type of resource | text |
|---|---|
| Form | electronic resource; remote; computer; online resource |
| Extent | 1 online resource. |
| Place | California |
| Place | [Stanford, California] |
| Publisher | [Stanford University] |
| Copyright date | 2021; ©2021 |
| Publication date | 2021; 2021 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Author | Anand, Namrata |
|---|---|
| Degree supervisor | Huang, Possu |
| Thesis advisor | Huang, Possu |
| Thesis advisor | Das, Rhiju |
| Thesis advisor | Fordyce, Polly |
| Thesis advisor | Kundaje, Anshul, 1980- |
| Degree committee member | Das, Rhiju |
| Degree committee member | Fordyce, Polly |
| Degree committee member | Kundaje, Anshul, 1980- |
| Associated with | Stanford University, Department of Bioengineering |
Subjects
| Genre | Theses |
|---|---|
| Genre | Text |
Bibliographic information
| Statement of responsibility | Namrata Anand-Achim. |
|---|---|
| Note | Submitted to the Department of Bioengineering. |
| Thesis | Thesis Ph.D. Stanford University 2021. |
| Location | https://purl.stanford.edu/kg677jf0057 |
Access conditions
- Copyright
- © 2021 by Namrata Anand
Also listed in
Loading usage metrics...