Resource-efficient quantum algorithm for protein folding

Predicting the three-dimensional (3D) structure of a protein from its primary sequence amino acids is known as folding (PF) problem. Due to central role proteins' 3D structures in chemistry, biology and medicine applications (e.g., drug discovery) this subject has been intensively studied for over half century. Although classical algorithms provide practical solutions, sampling conformation space small proteins, they cannot tackle intrinsic NP-hard complexity problem, even reduced simplest Hydrophobic-Polar model. While fault-tolerant quantum computers are still beyond reach state-of-the-art technologies, there evidence that can be successfully used on Noisy Intermediate-Scale Quantum (NISQ) accelerate energy optimization frustrated systems. In work, we present model Hamiltonian with $\mathcal{O}(N^4)$ scaling corresponding variational algorithm polymer chain $N$ monomers tetrahedral lattice. The reflects many physico-chemical properties protein, reducing gap between coarse-grained representations mere lattice models. We use robust versatile optimisation scheme, bringing together specifically adapted cost functions evolutionary strategies (genetic algorithms), simulate 10 acid Angiotensin peptide 22 qubits. same method also applied study 7 neuropeptide using 9 qubits an IBM Q 20-qubit computer. Bringing recent advances building gate-based noise-tolerant hybrid quantum-classical algorithms, work paves way towards accessible relevant scientific experiments real processors.