Quantum IsoRank: Efficient Alignment of Multiple PPI Networks

TL;DR

This paper introduces a quantum computing method for aligning multiple protein-protein interaction networks, leveraging quantum phase estimation to improve efficiency over classical algorithms in biological network analysis.

Contribution

It presents a novel quantum algorithm based on phase estimation for network alignment, outperforming classical methods in computational complexity.

Findings

Quantum approach achieves faster alignment computation.

Uses quantum phase estimation to find principal eigenvector.

Outperforms classical algorithms in efficiency.

Abstract

Comparative analyses of protein-protein interaction networks play important roles in the understanding of biological processes. However, the growing enormity of available data on the networks becomes a computational challenge for the conventional alignment algorithms. Quantum algorithms generally provide greater efficiency over their classical counterparts in solving various problems. One of such algorithms is the quantum phase estimation algorithm which generates the principal eigenvector of a stochastic matrix with probability one. Using the quantum phase estimation algorithm, we introduce a quantum computing approach for the alignment of protein-protein interaction networks by following the classical algorithm IsoRank which uses the principal eigenvector of the stochastic matrix representing the Kronecker product of the normalized adjacency matrices of networks for the pairwise alignment. We also present a greedy quantum measurement scheme to efficiently procure the alignment from the output state of the phase estimation algorithm where the eigenvector is encoded as the amplitudes of this state. The complexity of the quantum approach outperforms the classical running time.