ProQSAR: A modular and reproducible framework for small-data QSAR modeling with fit-and-use models
Tuyet-Minh Phan, Tieu-Long Phan, Tieu-Long Phan, Tieu-Long Phan, Phuoc-Chung Van Nguyen, Hoang-Son Lai Le, Van-Thinh To, Tuyen Ngoc Truong, Daniel M. Merkle, Peter Florian Stadler · Journal of Cheminformatics · 2026
Abstract Background Quantitative structure-activity relationship (QSAR) models are central to computer-aided drug discovery and predictive toxicology, but practical adoption is often impeded by ad-hoc tooling, inconsistent validation protocols, and poor reproducibility. Objective We introduce , a modular, reproducible workbench that formalizes end-to-end QSAR development while permitting independent use of each component. Methods composes interchangeable modules for standardization, feature generation, splitting (including scaffold- and cluster-aware splits), preprocessing, outlier handling, scaling, feature selection, model training and tuning, statistical comparison, conformal calibration, and applicability-domain assessment. The pipeline can run end-to-end to produce versioned artifact bundles (serialized models) and analyst-oriented reports suitable for deployment and audit. Results On representative benchmarks evaluated under Bemis–Murcko scaffold split, attains state-of-the-art descriptor-based performance: the lowest mean RMSE across the regression suite (, , ; mean RMSE $$0.658\pm 0.11$$ 0.658 ± 0.11 ), including a substantial improvement on (RMSE $$0.494$$ 0.494 vs. $$0.731$$ 0.731 for a leading graph method). On quantum mechanical benchmarks, demonstrated superior performance on the single-task dataset and maintained competitive results on the multi-task dataset. For classification, achieves the top ROC–AUC on (91.4%) while remaining competitive across other benchmark (overall classification average $$70.4\pm 11.6$$ 70.4 ± 11.6 ). Crucially, all predictions are accompanied by cross-conformal prediction and explicit applicability-domain flags that identify out-of-distribution entries, enabling calibrated and decision support. Availability is released on , , and ; all releases embed full provenance (parameters, package versions, checksums) to ensure reproducibility. Scientific contribution (i) enforces best-practice, group-aware validation together with formal statistical comparisons across models, (ii) integrates calibrated uncertainty quantification (cross-conformal prediction) and applicability-domain diagnostics for interpretable, risk-aware predictions, and (iii) exposes both a composable developer API and a one-click pipeline that generates deployment-ready artifacts and human-readable reports, demonstrated on representative benchmarks.