Robust Quantum Reservoir Computing for Molecular Property Prediction

TL;DR

This paper introduces a robust quantum reservoir computing approach for predicting molecular properties, demonstrating improved performance on small datasets and enhanced interpretability of quantum embeddings in biomedical drug discovery.

Contribution

The study applies quantum reservoir computing to molecular property prediction, showing advantages over classical models especially with limited data and providing insights into quantum feature transformations.

Findings

QRC outperforms classical models on small datasets

Quantum embeddings are more interpretable in reduced dimensions

QRC offers a promising quantum approach for drug discovery

Abstract

Machine learning has been increasingly utilized in the field of biomedical research to accelerate the drug discovery process. In recent years, the emergence of quantum computing has been followed by extensive exploration of quantum machine learning algorithms. Quantum variational machine learning algorithms are currently the most prevalent but face issues with trainability due to vanishing gradients. An emerging alternative is the quantum reservoir computing (QRC) approach, in which the quantum algorithm does not require gradient evaluation on quantum hardware. Motivated by the potential advantages of the QRC method, we apply it to predict the biological activity of potential drug molecules based on molecular descriptors. We observe more robust QRC performance as the size of the dataset decreases, compared to standard classical models, a quality of potential interest for pharmaceutical datasets of limited size. In addition, we leverage the uniform manifold approximation and projection technique to analyze structural changes as classical features are transformed through quantum dynamics and find that quantum reservoir embeddings appear to be more interpretable in lower dimensions.