Gate-based Quantum Computing for Protein Design

TL;DR

This paper explores using gate-based quantum computing, specifically Grover's algorithm, to efficiently address the protein design problem by finding low-energy amino acid configurations faster than classical methods.

Contribution

It introduces a quantum circuit approach utilizing Grover's algorithm for protein design, demonstrating quadratic speedup and implementation on simulators and real quantum devices.

Findings

Quantum circuits achieved quadratic speedup in protein design search.

Simulations confirmed the effectiveness of the quantum approach.

Initial experiments on real quantum devices demonstrated feasibility.

Abstract

Protein design is a technique to engineer proteins by modifying their sequence to obtain novel functionalities. In this method, amino acids in the sequence are permutated to find the low energy states satisfying the configuration. However, exploring all possible combinations of amino acids is generally impossible to achieve on conventional computers due to the exponential growth of possibilities with the number of designable sites. Thus, sampling methods are currently used as a conventional approach to address the protein design problems. Recently, quantum computation methods have shown the potential to solve similar types of problems. In the present work, we use the general idea of Grover's algorithm, a pure quantum computation method, to design circuits at the gate-based level and address the protein design problem. In our quantum algorithms, we use custom pair-wise energy tables consisting of eight different amino acids. Also, the distance reciprocals between designable sites are included in calculating energies in the circuits. Due to the noisy state of current quantum computers, we mainly use quantum computer simulators for this study. However, a very simple version of our circuits is implemented on real quantum devices to examine their capabilities to run these algorithms. Our results show that using $\mathcal{O}(\sqrt N)$ iterations, the circuits find the correct results among all $N$ possibilities, providing the expected quadratic speed up of Grover's algorithm over classical methods.