The prediction of molecular properties from SMILES strings is a cornerstone of modern drug discovery and materials science, yet traditional computational methods often face a trade-off between accuracy and speed. This paper presents a comprehensive hybrid quantum-classical architecture that synergistically combines a Bidirectional Long Short-Term Memory (BiLSTM) network with a variational quantum circuit (VQC) to perform multi-task regression on molecular properties. Our system introduces a novel quantum feature layer, implemented in PennyLane and integrated with PyTorch, which employs data reuploading and configurable entanglement patterns (linear, full, pairwise) to generate expressive quantum descriptors. These descriptors are concatenated with classical features and fed into a regression head, enabling the simultaneous prediction of eight key molecular properties. We validate our approach through rigorous experimentation on benchmark datasets (ESOL, FreeSolv, QM9) and demonstrate that the hybrid model consistently outperforms a strong classical BiLSTM baseline, achieving an $\mathbf{R}^{2}$ of up to 0.94 on QM9 and reducing inference time to seconds per molecule. Furthermore, we detail a complete production-ready pipeline, including SMILES tokenization, exploratory data analysis (EDA), handling of invalid inputs, and multiple deployment scenarios such as REST API integration for LIMS/ERP systems, active learning loops, and large-scale batch processing on cloud infrastructure. This work establishes a scalable and robust framework for integrating quantum computing into real-world cheminformatics workflows, offering a significant step toward quantumaccelerated drug design.
The prediction of molecular properties from SMILES strings is a cornerstone of modern drug discovery and materials science, yet traditional computational methods often face a trade-off between accuracy and speed. This paper presents a comprehensive hybrid quantum-classical architecture that synergistically combines a Bidirectional Long Short-Term Memory (BiLSTM) network with a variational quantum circuit (VQC) to perform multi-task regression on molecular properties. Our system introduces a novel quantum feature layer, implemented in PennyLane and integrated with PyTorch, which employs data reuploading and configurable entanglement patterns (linear, full, pairwise) to generate expressive quantum descriptors. These descriptors are concatenated with classical features and fed into a regression head, enabling the simultaneous prediction of eight key molecular properties. We validate our approach through rigorous experimentation on benchmark datasets (ESOL, FreeSolv, QM9) and demonstrate that the hybrid model consistently outperforms a strong classical BiLSTM baseline, achieving an R2 of up to 0.94 on QM9 and reducing inference time to seconds per molecule. Furthermore, we detail a complete production-ready pipeline, including SMILES tokenization, exploratory data analysis (EDA), handling of invalid inputs, and multiple deployment scenarios such as REST API integration for LIMS/ERP systems, active learning loops, and large-scale batch processing on cloud infrastructure. This work establishes a scalable and robust framework for integrating quantum computing into real-world cheminformatics workflows, offering a significant step toward quantumaccelerated drug design.
Classical machine learning models and quantum kernel methods often struggle to capture quantum-coherent molecular features under the constraints of noisy intermediate-scale quantum (NISQ) hardware, limiting both predictive accuracy and scalability. This paper introduces the Molecular Quantum Particle Algorithm (MQPA), a hybrid quantum–classical framework designed to achieve chemically accurate property prediction by integrating handcrafted molecular descriptors with parameterized quantum circuits. Molecular inputs, expressed as SMILES strings, are processed via RDKit and encoded through angle-based quantum gates with entangling layers in Qiskit [1]. Quantum parameters are optimized using simultaneous perturbation stochastic approximation (SPSA) [2], while classical regression layers leverage Adam [3] with Low-Rank Adaptation (LoRA) fine-tuning [4]. Across 38 systematically varied subsets of the QM9 benchmark [5,6], MQPA achieves a mean absolute error (MAE) of 0.03 kcal/mol, outperforming both deep learning and quantum kernel baselines by 41% and surpassing the chemical accuracy threshold of 0.1 kcal/mol [7–9]. Extensive cross-platform benchmarking—including IBM Quantum, Qiskit Aer, and NVIDIA H100 clusters—demonstrates strong reproducibility, with inter-backend MAE variance below 3%. Positioned as a scalable and hardware-aware solution, MQPA offers practical advantages for quantum-enhanced molecular property prediction across applications in quantum chemistry, drug discovery, materials design, and environmental modeling.
Finding drug-like molecules is a time- and money-consuming task that requires virtual screening techniques capable of accurately representing non-linear intermolecular relationships. Classical ML models usually consider molecular graphs or SMILES sequences, but cannot jointly exploit both structural and sequential information, let alone simultaneously. Of greater importance, numerous deep learning methods are hindered by poor generalisation, black-box predictions, and inadequate interpretability, reducing their feasibility for application in real-life drug discovery pipelines. In this paper, we present QDrugDiscoverAI, a hybrid quantum–classical deep learning framework for predicting molecular properties and optimising molecules to overcome these challenges. We present a framework for multimodal molecular representation learning that combines graph attention networks and transformer-based sequence encoders with variational quantum circuits to increase expressive capacity within a single architecture. It implements a dual-stream representation-learning mechanism with a multimodal feature fusion layer. It includes an explainability module that uses attention weights and SHAP values to identify substructures that contribute to model predictions and enhance interpretability. Moreover, QDrugDiscoverAI also includes a quantum–assisted reinforcement-learning module for a molecule-optimisation workflow, enabling more efficient exploration of the chemical design space while leveraging the controllability of quantum simulation. The experiments on three benchmark datasets, i.e., Lipophilicity, HIV, and BBBP, show improvements of 12.4%, 9.6%, and 11.1% in accuracy, AUC, and F1-score, respectively, over the matched classical and unimodal baselines. Quantitative evaluation of quantum components and fusion strategies is also conducted through ablation studies to demonstrate their roles further. In summary, the framework proposed serves as a systematic integration of interpretable hybrid learning, showing one way it could be organised to become increasingly efficient, and demonstrating the advantage of using quantum-simulated components to enhance classical deep learning models for exploration and predictive performance in drug discovery workflows. The implementation code and supporting scripts for the proposed QDrugDiscoverAI framework are publicly available at: https://github.com/Sripada-RamaSree/QuantumDrugAI
The pharmaceutical industry faces mounting challenges in drug discovery, where exploring a chemical space exceeding 10 60 possible compounds remains computationally prohibitive. Traditional in-silico screening approaches struggle to capture the quantum mechanical interactions that determine molecular behavior. To overcome these limitations, this study introduces a hybrid quantum–classical machine learning framework that uniquely integrates molecular graph theory with quantum-enhanced learning for accelerated drug discovery. Using the QM9 benchmark dataset containing 133,885 molecules of low-to-medium structural complexity, we extract Laplacian eigenvalue spectra as topological descriptors that preserve both connectivity and electronic information. These features are jointly processed by Random Forest (RF), Support Vector Machine (SVM), and a Variational Quantum Classifier (VQC) in a weighted ensemble architecture designed for balanced generalization. The Quantum Graph Neural Network (QGNN) component achieves 97.5% accuracy, while the hybrid ensemble attains 100.0% accuracy, confirming strong predictive reliability. Although quantum training requires roughly 3.2× longer computation, it captures high-dimensional correlations that classical methods overlook. Overall, the proposed framework bridges the gap between theoretical quantum computing and practical pharmaceutical modeling, providing a scalable, physics-inspired pathway toward intelligent molecular screening and next-generation drug discovery systems.
This work aims to propose a Quantum-Classical Neural Network with Variational quantum circuit (QCNN-VQC) to improve the accuracy of molecular docking in drug discovery. The model combines the classical feature extraction framework with quantum representations and energy-based optimisation to simulate intricate protein-ligand interactions. The molecular features are initially encoded into quantum states by first a standard neural network and then by encoding them into quantum states, where non-linear correlations and entanglement effects are modeled by variational circuits. An optimization strategy is used, that is quantum-inspired to achieve the minimum docking energy and find stable binding conformations. The experimental analysis of molecular docking datasets reveals that the proposed model achieves an accuracy of 91.2% compared to the classical neural networks (78.5%), and deep learning models (82.3%). The model also has a lower RMSE of 0.38 and higher R2score of 0.89 which means that it predicts better. The results of energy minimization indicate that the docking energy is greatly reduced, which proves the increased binding stability. The hybrid model is slightly more expensive to compute but it has competitive inference time. In general, QCNN VQC architecture offers a scalable and efficient drug discovery system; it can be optimally applied to quantum optimization in conjunction with deep learning and can result in a better prediction and strong molecular interaction modeling.
Hybrid quantum-classical neural networks (HQCNNs) represent a promising but largely unproven paradigm for enhancing molecular property prediction in drug discovery. Despite growing interest and optimistic claims in recent literature, rigorous benchmarks comparing these hybrid architectures against strong classical baselines remain scarce. This work implements and systematically evaluates a representative HQCNN architecture against its classical equivalent using standardized molecular datasets (QM9 and ZINC-250K) within a reproducible, multi-seed experimental framework. Contrary to optimistic expectations, our results reveal that the hybrid model consistently underperforms the classical baseline, showing a 49% increase in MAE on QM9 (1.0834 vs 0.7256) and a devastating 107% degradation on ZINC (1.4006 vs 0.6755), with R2values dropping by over 80% on both datasets. Through detailed analysis, we identify fundamental challenges including information bottlenecks during quantum encoding and difficult optimization landscapes in variational quantum circuits as primary causes of performance degradation. This work provides a crucial, datadriven reality check for the field, demonstrating that current “plug-and-play” quantum layer approaches may be insufficient for practical quantum advantages. All experimental code and data are publicly released to enable community replication and investigation of more sophisticated quantum-native architectures.
Predicting drug-target binding affinity with limited training data remains a central challenge in computational drug discovery. We introduce a hybrid quantum-classical framework combining molecular descriptors with variational quantum circuits for interpretable binding affinity prediction. Seven physicochemical descriptors (MW, logP, TPSA, HBD/HBA, rotatable bonds, aromatic rings) were encoded into 6-qubit variational circuits using parameterized rotations and controlled-Z entanglement. Quantum kernels were extracted and used for regression on 1,200 BindingDB molecules across kinase, GPCR, and protease targets, with external validation on 300 ChEMBL compounds. Performance was compared against SVR, random forest, gradient boosting, and neural network baselines using 100-fold bootstrap validation. Variational quantum regression achieved MSE = 0.056 ± 0.009, outperforming SVR by 32% (0.082 ± 0.013), RF by 28%, and GBM by 24%. Quantum advantage was most significant in low-data regimes (n < 500), where VQR maintained R2 > 0.85 versus R2 < 0.72 for classical methods. External validation confirmed generalization (R2 = 0.83). Our Explainable Quantum Pharmacology framework revealed TPSA and logP as dominant predictive features through gradientbased sensitivity analysis, aligning with established medicinal chemistry principles. Quantum kernels provide measurable, chemically interpretable improvements for molecular property prediction in data-limited scenarios. This work demonstrates practical quantum advantage for early-stage drug discovery applications on near-term quantum hardware.