We present the first systematic, hardware-executed benchmark of twelve distinct quantum data-encoding strategies for drug-response prediction on a real superconducting quantum processing unit (QPU). All experiments were conducted on the IQM Garnet 20-qubit QPU via the IQM Resonance cloud platform, using the Qrisp quantum-software framework (v 0.8.2). Each encoding was evaluated on n = 50 stratified samples drawn from the Genomics of Drug Sensitivity in Cancer dataset (GDSC2, 242 036 drug–cell-line pairs), targeting the natural-log IC 50 response variable. Variational weights were optimised offline with the gradient-free COBYLA algorithm before hardware submission. Every circuit was executed with 1024 shots; the regression signal is the zero-qubit Pauli expectation value ⟨ Z 0 ⟩. Results show that the QAOA-inspired encoding achieves the best RMSE of 3.314 and is statistically superior ( p < 0.05, Wilcoxon signed-rank test) to six of the remaining eleven encodings. Hardware-efficient entanglement structures—specifically alternating cost and mixer layers—provide a systematic advantage over purely rotational or diagonal encodings under realistic noise conditions. This work constitutes a reproducible baseline for noise-aware quantum machine learning on pharmaceutical data; all code, data, and raw QPU outputs are publicly released.
We present the first systematic, hardware-executed benchmark of twelve distinct quantum data-encoding strategies for drug-response prediction on a real superconducting quantum processing unit (QPU). All experiments were conducted on the IQM Garnet 20-qubit QPU via the IQM Resonance cloud platform, using the Qrisp quantum-software framework (v 0.8.2). Each encoding was evaluated on n = 50 stratified samples drawn from the Genomics of Drug Sensitivity in Cancer dataset (GDSC2, 242 036 drug–cell-line pairs), targeting the natural-log IC 50 response variable. Variational weights were optimised offline with the gradient-free COBYLA algorithm before hardware submission. Every circuit was executed with 1024 shots; the regression signal is the zero-qubit Pauli expectation value ⟨ Z 0 ⟩. Results show that the QAOA-inspired encoding achieves the best RMSE of 3.314 and is statistically superior ( p < 0.05, Wilcoxon signed-rank test) to six of the remaining eleven encodings. Hardware-efficient entanglement structures—specifically alternating cost and mixer layers—provide a systematic advantage over purely rotational or diagonal encodings under realistic noise conditions. This work constitutes a reproducible baseline for noise-aware quantum machine learning on pharmaceutical data; all code, data, and raw QPU outputs are publicly released.
Abstract Quantum machine learning for genomic sequence classification faces three compounding constraints on current hardware: barren plateaus in variational training, exponential concentration of quantum kernel expressivity in highdimensional feature spaces, and noise accumulation that limits practical circuit depth to fewer than 50 layers on current NISQ devices. Existing quantum genomics studies address these constraints using fixed-architecture circuits without noise-calibrated validation on standardized benchmarks, leaving the potential of adaptive circuit topology largely unexplored. Three QAOA variants are developed and systematically compared to address this gap. Classical QAOA establishes a baseline using fixed p-layer circuits with standard phase and mixing operators. Alternating QAOA introduces structured CNOT entanglement between alternating qubit pairs to evaluate whether non-local correlations improve genomic pattern recognition. The primary contribution, Adaptive QAOA, employs gradient-driven operator selection from feature-weighted Pauli operator pools, generating variable-depth problemspecific circuits that adapt their topology to dataset structure rather than applying universal ans¨atze. All three architectures are evaluated through quantum simulation on two standardized genomic benchmarks: Human Nontata Promoters (36,131 sequences) and Demo Coding vs. Intergenic Sequences (100,000 sequences). Experiments are conducted under both noiseless conditions and noise models calibrated to IBM superconducting hardware specifications, across feature space dimensionalities ranging from 2 to 10 dimensions using 10-fold cross-validation. Adaptive QAOA consistently achieves the highest precision (approximately 75% under noiseless simulation) and maintains perfect specificity across both datasets and noise conditions, results not observed for fixed-depth variants. These performance gains are accompanied by reduced hardware resource consumption, with Adaptive QAOA requiring the fewest qubits and lowest quantum usage time (∼0.52–0.60 s) across most configurations. Superior coherence characteristics — average T1 of 82.12 μs and T2 of 55.45 μs, exceeding Classical QAOA by 13.6% and 113.3% respectively — provide a physical basis for reduced error accumulation under realistic noise constraints. All QAOA variants maintain stable classification performance across all tested dimensionalities, contrasting with compounding errors documented for fixed-architecture quantum kernel methods on high-dimensional genomic data. These results establish that gradient-driven adaptive circuit topology with domain-specific feature weighting represents a more effective design strategy than uniform depth scaling or entanglement-based ansatz modification for near-term quantum genomic classification.
This work presents a hardware-validated quantum framework for encoding cancer gene expression data and evaluating its alignment with drug response signals. Using a quantum kernel-based architecture (Qube Engine), classical gene expression data from CCLE is transformed into quantum states and executed on IBM Quantum hardware. Outputs are compared against IC50 drug response values from the GDSC dataset. The system demonstrates strong stability under real hardware noise (~97.15%) and maintains measurable biological relevance, achieving a correlation of ~0.45 with IC50 values. This work highlights the feasibility of applying quantum machine learning techniques to real-world biomedical datasets under NISQ constraints.
Large compound libraries are reduced to candidate’s worth testing by computational prediction of drug-target interactions (DTI). We investigate whether this effort is aided by variational quantum circuits appropriate for noisy intermediate-scale quantum (NISQ) devices. Our hybrid model uses classical networks to encode ECFP4 fingerprints and protein sequence descriptors, compresses them to eight rotation angles, and then processes them using a four-layer hardware-efficient circuit that is read out using local Pauli-Z expectations. Realistic noise models are used to simulate the circuits, and it is trained end-to-end using gradients that match the parameter-shift algorithm. The model achieved test AUROCs of 0.896, 0.892, 0.911, and 0.878 on Davis, KIBA, BindingDB, and BIOSNAP. These results outperform a kernel SVM and logistic regression, although they lag behind a graph network on three datasets, a deep network (0.915–0.927), and gradient-boosted trees. Removing entanglement had little effect on the performance of a parameter-matched classical head, and no low-data advantage emerged. Hardware tolerance was demonstrated by the circuit itself. Because ranking is invariant to uniform noise attenuation, local readout avoided barren plateaus, 256 shots were sufficient, and 1% two-qubit error cost only 0.0025 AUROC. Calibration was recovered by recalibrating the reading. Hardware training costs, not noise, are the binding restriction. For drug-screening decisions, these predictions provide an evidence-based ranking of candidate compound–target pairs for experimental follow-up. The comparison with classical methods indicates whether adding a quantum circuit is justified by the resulting predictive performance and computational cost.
Quantum machine learning (QML) has generated significant excitement as a potential game-changer in artificial intelligence. At the heart of any QML model lies a fundamental challenge: how do we translate classical data into the language of quantum mechanics? This encoding process is crucial because the way we represent information in a quantum system can make or break model performance. While researchers have proposed several encoding strategies over the years, most studies focus on idealized conditions, leaving a significant gap in our understanding of how these methods actually behave under the noisy, imperfect conditions of real quantum hardware. In this work, we address this gap through a systematic, simulation-based comparison of three widely used encoding strategies—Angle Encoding, Amplitude Encoding, and the entangling ZZFeatureMap—within a variational quantum classification framework. We evaluate these approaches on standard benchmark datasets under both noiseless and noisy conditions that mimic real quantum devices. To ensure our findings are statistically reliable, each noisy experiment is repeated across 10 independent runs, with results reported using mean accuracy, standard deviation, and 95% confidence intervals. Our results reveal a nuanced but sobering story. While the entangling ZZFeatureMap outperforms simpler methods in ideal noise-free settings, these advantages largely vanish under realistic depolarizing noise. When statistically averaged, all encoding strategies converge to near-random classification accuracy under moderate noise levels. These findings highlight the fragility of quantum advantages on today's noisy intermediate-scale quantum (NISQ) hardware and emphasize the critical need for noise-aware, statistically grounded evaluation protocols when assessing the practical viability of QML models.
Quantum-classical Hybrid Machine Learning (QHML) models are recognized for their robust performance and high generalization ability even for relatively small datasets. These qualities offer unique advantages for anti-cancer drug response prediction, where the number of available samples is typically small. However, such hybrid models appear to be very sensitive to the data encoding used at the interface of a neural network and a quantum circuit, with suboptimal choices leading to stability issues. To address this problem, we propose a novel strategy that uses a normalization function based on a moderated gradient version of the tanh. This method transforms the outputs of the neural networks without concentrating them at the extreme value ranges. Our idea was evaluated on a dataset of gene expression and drug response measurements for various cancer cell lines, where we compared the prediction performance of a classical deep learning model and several QHML models. These results confirmed that QHML performed better than the classical models when data was optimally normalized. This study opens up new possibilities for biomedical data analysis using quantum computers.
The continuous progress of quantum technologies has spurred the exploration of their potential applications across diverse fields, particularly in combinatorial optimisation. In this work, we study a quantum-classical optimisation framework based on Pauli Correlation Encoding, an encoding scheme that can represent m binary variables using a polynomial number of qubits. To evaluate the performance of the method, we use three classical optimisation problems against the instances of the QOPTLib benchmark. The study includes an analysis of the impact of the compression order of the encoding scheme, the problem structure, and hyperparameter selection on solution quality, as well as the role of post-processing in improving performance. Additionally, we study the effect of shot-based execution and hardware noise, showing how these factors influence both the accuracy of expected value estimation and the overall dynamics of the optimisation process. The results indicate that the proposed PCE-based framework achieves competitive performance against the benchmark and, in several cases, obtains equivalent or even superior solutions, highlighting its potential as an efficient encoding strategy for quantum optimisation in the NISQ and near fault-tolerant era.