Evaluating Effects of Augmented SELFIES for Molecular Understanding Using QK-LSTM

TL;DR

The paper addresses the challenge of efficiently predicting molecular properties and identifying adverse effects using a hybrid quantum-classical framework, QK-LSTM, which injects quantum Kernel feature maps into LSTM models. It also investigates the impact of augmenting molecular representations with SELFIES versus SMILES, showing that SELFIES augmentation yields significant improvements in both classical and quantum-classical settings. The study demonstrates that quantum-classical hybrids can achieve performance comparable to classical models on cheminformatics tasks, with augmentation serving as a practical lever for improved generalization. These findings highlight the potential of quantum-enhanced sequence learning in drug discovery and motivate further research with larger datasets and advancing quantum hardware.

Abstract

Identifying molecular properties, including side effects, is a critical yet time-consuming step in drug development. Failing to detect these side effects before regulatory submission can result in significant financial losses and production delays, and overlooking them during the regulatory review can lead to catastrophic consequences. This challenge presents an opportunity for innovative machine learning approaches, particularly hybrid quantum-classical models like the Quantum Kernel-Based Long Short-Term Memory (QK-LSTM) network. The QK-LSTM integrates quantum kernel functions into the classical LSTM framework, enabling the capture of complex, non-linear patterns in sequential data. By mapping input data into a high-dimensional quantum feature space, the QK-LSTM model reduces the need for large parameter sets, allowing for model compression without sacrificing accuracy in sequence-based tasks. Recent advancements have been made in the classical domain using augmented variations of the Simplified Molecular Line-Entry System (SMILES). However, to the best of our knowledge, no research has explored the impact of augmented SMILES in the quantum domain, nor the role of augmented Self-Referencing Embedded Strings (SELFIES) in either classical or hybrid quantum-classical settings. This study presents the first analysis of these approaches, providing novel insights into their potential for enhancing molecular property prediction and side effect identification. Results reveal that augmenting SELFIES yields in statistically significant improvements from SMILES by a 5.97% improvement for the classical domain and a 5.91% improvement for the hybrid quantum-classical domain.

Figures (4)

• Figure 1: Sample benzene molecule that relates to both the SMILES and SELFIES samples supplied
• Figure 2: QK-LSTM architecture used for training; ($C_{t-1}, C_t$) represent the cell state, ($X_{t}$) represents the input, ($h_{t-1}, h_t$) represent the output state. The $VQC_1$ wire represents the forget gate, deciding if the input should be added to memory. The ($VQC_2, VQC_3$) wires represent the update gate, updating the cell memory if needed. The $VQC_4$ wire represents the output gate, outputting the result of the QK-LSTM to the rest of the model.
• Figure 3: QK-LSTM basic entangler circuit; (a) angle encoding converts classical features ($f_1, f_4, f_3, f_4$) to quantum states, (b) parametric quantum circuit entangles quantum states, (c) qubits measured, and bloated to original higher-dimension space. bergholm2018pennylane.
• Figure 4: Overview of the LSTM process.