arXiv:2610.00887
A Gate-Based Quantum Computing Framework for Codon Optimization
Fatemeh Ghasemi, Kion Kim
Abstract
Codon optimization is a challenging combinatorial optimization problem with important applications in synthetic biology, protein expression, and biotechnology. While quantum annealing has previously been explored for this problem, the application of gate-based quantum algorithms remains largely unexplored. In this work, we present a gate-based quantum computing framework for codon optimization by formulating the optimization objective as an Ising Hamiltonian and expressing it in terms of Pauli operators suitable for gate-based quantum processors. The proposed framework is investigated using the Variational Quantum Eigensolver (VQE), Sampling Variational Quantum Eigensolver (SVQE), and the Quantum Approximate Optimization Algorithm (QAOA), with benchmarking against exact diagonalization, classical optimization, and quantum annealing. Using a three-amino-acid benchmark, the proposed implementations successfully reproduce the exact ground-state solutions while identifying optimized ansatz selections, optimization strategies, initialization schemes, and hardware execution settings that provide reproducible performance on current noisy quantum hardware. Hardware demonstrations are performed on IBM Quantum processors to evaluate the practical implementation of the proposed workflow. As a work in progress, this study establishes a reproducible computational framework for gate-based quantum codon optimization and provides the foundation for future investigations of larger protein sequences as quantum hardware and algorithms continue to advance.
AI Summary
The paper investigates how gate-based variational quantum algorithms can implement a codon-optimization objective previously used with quantum annealing. A binary formulation combines codon usage, target GC content, repeated-nucleotide penalties, and one-hot codon-selection constraints. Applying $q_i=(1-Z_i)/2$ produces an Ising cost Hamiltonian for VQE, sampling VQE (SVQE), and QAOA. The demonstration uses Glu–Gly–Pro, with 10 candidate codons encoded in 10 qubits and only 32 valid coding sequences. Codon-usage and repeated-nucleotide terms are disabled, leaving a 50% GC-content target and selection penalties. Exact enumeration identifies four degenerate ground states with reported energy $-6.2469$. Classical optimization, quantum annealing, simulated SVQE, and simulated single-layer QAOA recover all four. VQE performance is described inconsistently: the results report run dependence, while the implementation section reports consistent ground-energy recovery for the selected configuration. IBM hardware experiments test SVQE with multiple random initializations and QAOA with fixed parameters $(\beta,\gamma)=(0.5,0.8)$. Ten SVQE initializations recover all four optima, using approximately 11,615 seconds of QPU time. QAOA recovers three optima with 4,096 shots, and all four with 6,000 or 10,000 shots, or with dynamical decoupling at 4,096 shots. Reported individual optimal-string probabilities are generally around $10^{-4}$–$10^{-2}$. The study is explicitly a small proof of concept, not a demonstration of quantum superiority. Scaling to longer sequences and comprehensive performance assessment are deferred; hardware statistics are representative executions rather than averages over repeated trials.
Central claim may not hold: The paper does not reconcile its report of inconsistent VQE ground-state recovery with its assertion that the selected VQE configuration consistently reproduced the exact ground-state energy. Table I's caption and the Conclusions claim that all investigated methods recovered all four ground states, although no such VQE results are provided and Results reports inconsistent VQE recovery. The claimed reproducible hardware performance is supported by representative executions and qualitative statements about repeats, without repeated-run success statistics establishing reliable recovery of the full ground-state set.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.