Accelerating De Novo Genome Assembly via Quantum-Assisted Graph Optimization with Bitstring Recovery

Genome sequencing is essential to decode genetic information, identify organisms, understand diseases and advance personalized medicine. A critical step in any genome sequencing technique is genome assembly. However, de novo genome assembly, which involves constructing an entire genome sequence from scratch without a reference genome, presents significant challenges due to its high computational complexity, affecting both time and accuracy. In this study, we propose a hybrid approach utilizing a quantum computing-based optimization algorithm integrated with classical pre-processing to expedite the genome assembly process. Specifically, we present a method to solve the Hamiltonian and Eulerian paths within the genome assembly graph using gate-based quantum computing through a Higher-Order Binary Optimization (HOBO) formulation with the Variational Quantum Eigensolver algorithm (VQE), in addition to a novel bitstring recovery mechanism to improve optimizer traversal of the solution space. A comparative analysis with classical optimization techniques was performed to assess the effectiveness of our quantum-based approach in genome assembly. The results indicate that, as quantum hardware continues to evolve and noise levels diminish, our formulation holds a significant potential to accelerate genome sequencing by offering faster and more accurate solutions to the complex challenges in genomic research.

💡 Research Summary

The paper presents a hybrid quantum‑classical framework aimed at accelerating de novo genome assembly, a computationally intensive task that requires solving Hamiltonian and Eulerian path problems on assembly graphs. The authors first perform extensive classical preprocessing of raw FASTQ reads, applying quality control, adapter trimming, and GC‑content analysis to obtain high‑quality short‑read and long‑read datasets. Short reads are transformed into de Bruijn graphs (DBG) where each (k‑1)‑mer forms a node and each k‑mer defines a directed edge, leading to an Eulerian path formulation. Long reads are aligned to construct Overlap‑Layout‑Consensus (OLC) graphs, with overlap scores serving as edge weights; the assembly problem becomes a Hamiltonian path search, which is NP‑complete.

To map these graph problems onto a quantum computer, the authors introduce a Higher‑Order Binary Optimization (HOBO) formulation. Unlike conventional QUBO approaches that rely on one‑hot encoding and require O(N²) binary variables (and thus a comparable number of qubits) for a graph with N vertices, the HOBO model encodes vertex order using a binary representation of size N·log₂N. This reduces qubit demand dramatically, making the approach more scalable on near‑term gate‑based quantum hardware.

The cost function is defined as the negative sum of weighted edge selections, where binary variables indicate whether two vertices appear consecutively in the path. Because the formulation includes higher‑order terms (typically up to fourth order), the authors convert it into a quadratic form compatible with Qiskit’s QuadraticProgram and then solve it using the Variational Quantum Eigensolver (VQE). The VQE circuit employs a hardware‑efficient ansatz (e.g., a layered rotation‑entanglement structure) and classical optimizers such as COBYLA or SPSA.

A novel “bitstring recovery” mechanism is added to mitigate measurement errors and the stochastic nature of VQE. After each VQE run, the measured bitstring may be partially correct; the recovery algorithm examines local correlations among bits and the gradient of the cost function to infer missing or flipped bits, reconstructs a more plausible candidate solution, and feeds it back into the optimizer for further refinement. This iterative correction improves both convergence speed and solution quality.

For benchmarking, the authors implement a classical dynamic‑programming (DP) algorithm with memoization that solves the Hamiltonian path problem exactly in O(N²·2^N) time but benefits from graph sparsity in practice. They compare DP results with those obtained from quantum simulations and from IBM quantum hardware (27‑qubit devices). Experiments on small graphs (4–6 nodes) demonstrate that the quantum approach, especially when combined with bitstring recovery, achieves costs within a few percent of the optimal DP solution. However, on larger graphs (≥10 nodes) the depth of the VQE circuit and hardware noise cause a degradation in performance, indicating that current devices are not yet sufficient for genome‑scale assemblies.

The paper concludes that the HOBO encoding substantially lowers qubit requirements, and the bitstring recovery scheme effectively combats noise, making the method promising for future fault‑tolerant quantum computers. The authors outline future work: optimizing HOBO‑to‑QUBO transformations, integrating error‑correction techniques, exploring deeper and more expressive ansätze, and scaling the approach to realistic genome sizes (hundreds of thousands to millions of nodes) using quantum‑accelerated hybrid algorithms. If quantum hardware continues to improve, this framework could provide faster and more accurate de novo genome assemblies, benefiting fields ranging from personalized medicine to evolutionary biology.