The ability to generate proteins with desired properties remains one of the most complex yet impactful problems in biology and medicine. In particular, peptides are small protein fragments that hold significant potential as therapeutics. Peptides have a $25 billion addressable market, and are known to “bridge the pharmaceutical gap” between small molecule drugs and larger biologics such as antibodies. However, many potential drug targets currently have no known peptide binders. This project aims to build a deep generative model that outputs potential peptide binders given the sequence and structure of any receptor protein. We use techniques from deep learning and natural language processing to create a sequence-to-sequence model based on protein primary structure. In addition, the model incorporates structural information from receptors to constrain peptides to a selected binding site. The overall goal of this project is to develop a faster and cheaper computational method for generating early-stage candidates of peptide therapeutics.

Team Members

[foreach 357]

• [if 397 not_equal=””][/if 397][395]

[/foreach 357]

Project Mentors, Sponsors, and Partners

• Tristan Bepler, PhD

• Timothy Lu, MD, PhD

• Jamie Spangler, PhD

• Faculty Mentor: Jessica Dunleavey, PhD

• TA Mentor: Sarah Lee, MS

Course Faculty

[foreach 429]

• [if 433 not_equal=””][/if 433][431]

[/foreach 429]

Project Links

Additional Project Information