Systems Biology Group Meeting

Lab meeting will take place on
Monday, November 16, 2020
10:00am ET

Presenter: Vladimir Gligorijevi, Ph.D., Flatiron Research Scientist, Systems Biology Group, CCB

Machine learning-guided design of proteins

Protein design has led to remarkable results in agriculture, medicine, and technology, including the development of new enzymes, peptides, and biomaterials. However, the protein design space arising from all possible combinations of amino acids is a large combinatorial space that is only sparsely functional. Thus, efficiently exploring the vast protein sequence space to reduce the burden on experimental approaches for protein design still remains a challenge.

In this talk, I will introduce our novel conditional generative model that combines denoising sequence-to-sequence autoencoders with a protein function classifier for generating protein sequences with specific functions. The proposed model can be used for accelerating sequence-based protein design by exploring the vast sequence space more effectively. I will discuss some ongoing work on using this method for designing novel metal-binding proteins as well as enzymes that catalyze decarboxylation.