Research

Protein molecules are polymers composed of 20 types of amino acid residues linked in a linear fashion, which fold into unique three-dimensional structures according to their amino acid sequences to carry out specific functions based on the folded structures. The total number of possible amino acid sequence patterns for a protein, with N being the number of amino acid residues, is an immense 20^N (for a protein with 100 residues, this is approximately 10^130 different patterns). It is believed that within this vast space of amino acid sequences, beyond the proteins that have evolved in nature, there exists many novel proteins that could be useful for medical, industrial, and cellular control design applications. We aim to explore this vast protein amino acid sequence space and create desired proteins from scratch by analyzing data of naturally occurring proteins, performing computer simulations, and conducting biochemical experiments.

We have discovered rules for designing proteins of various shapes by focusing on the backbone structures, such as secondary structure and loop lengths, rather than on the details of the amino acid sequences. Using these rules, we have successfully designed protein structures of various shapes with atomic-level precision. Interestingly, these proteins form extremely stable structures with denaturation temperatures above 100°C. Moving forward, we will further develop our protein design methods to create proteins with shapes that are entirely different from those of proteins found in nature and proteins with desired functions from the vast space of amino acid sequences.