Folding lattice proteins confined on minimal grids using a quantum-inspired encoding.

Area of Science:

• Computational biology
• Quantum computing
• Optimization problems

Background:

• Steric clashes complicate modeling dense protein systems with explicit-chain methods.
• Lattice protein energy minimization on a confined grid presents a complex optimization challenge.
• This problem shares similarities with scheduling problems and can be formulated as a QUBO.

Purpose of the Study:

• To investigate the efficacy of quadratic unconstrained binary optimization (QUBO) for solving lattice protein energy minimization.
• To compare classical and quantum-classical annealing approaches for this optimization problem.
• To benchmark QUBO-based methods against traditional programming techniques and exact enumeration.

Main Methods:

• Formulating the lattice protein energy minimization problem as a QUBO.
• Employing classical simulated annealing.
• Utilizing hybrid quantum-classical annealing on a D-Wave system.
• Testing linear and quadratic programming methods.

Main Results:

• The QUBO formulation was successfully solved for a lattice protein chain length of 48 using both simulated annealing and quantum-classical annealing.
• Hybrid quantum-classical annealing achieved solutions in approximately 10 seconds.
• Linear and quadratic programming methods showed limitations with protein chain constraints.

Conclusions:

• QUBO is a viable and efficient approach for solving dense protein system energy minimization problems.
• Quantum-classical annealing offers a swift and consistent method for tackling these complex optimization tasks.
• Further exploration of QUBO for protein modeling is warranted, especially for larger systems.