Structure-Guided Discovery of Submicromolar 1,2,4-Triazole–Schiff-Base Inhibitors of Glutathione Reductase In Vitro Data

This record provides the complete set of Python scripts used for the quantitative analysis of in vitro inhibition data reported in the study “Structure-Guided Discovery of Submicromolar 1,2,4-Triazole–Schiff-Base Inhibitors of Glutathione Reductase”. The scripts implement a reproducible pipeline for deriving absolute IC₅₀ values, assessing statistical uncertainty through bootstrap resampling, and generating interactive visualizations suitable for both supplementary material and independent validation. Scope and Objectives.The primary aim of these scripts is to ensure rigorous, transparent, and reproducible analysis of dose–response experiments for candidate glutathione reductase (GR) inhibitors. Each script corresponds to a specific compound or reference inhibitor (e.g., AUR-514 to AUR-518, quercetin) and is designed to operate on the raw percentage inhibition values obtained from the biochemical assays. Analytical Framework. Nonlinear regression: A four-parameter logistic (4-PL) model is fitted to raw inhibition data using SciPy’s curve_fit. The parameters bottom, top, IC₅₀, and Hill slope are explicitly estimated. Absolute IC₅₀ derivation: The midpoint is analytically determined only under the condition that 50% inhibition lies between the fitted asymptotes. Uncertainty quantification: A bootstrap protocol (5 seeds × 2,000 resamples = 10,000 fits per dataset) provides median values, interquartile ranges, and 95% confidence intervals. Distributional analysis: Kernel density estimation (KDE) of IC₅₀ replicates complements the descriptive statistics and reveals the reliability of the estimates. Visualization: Two interactive outputs are generated for each dataset: Dose–Response Curves with 95% CI bands IC₅₀ Distribution Histograms with KDE overlay Reproducibility.All scripts enforce deterministic random seeds, explicit bootstrap sample sizes, and clear output conventions. The outputs include interactive HTML visualizations and plain-text console summaries of parameter estimates. The workflow executes efficiently on standard CPU environments without GPU acceleration. Technical Requirements. Python ≥ 3.10 Core libraries: numpy, scipy, plotly No external dependencies or non-standard compilers are required. Availability and Citation.The scripts are made openly available to facilitate reproducibility and secondary analysis. Users are encouraged to cite both the Zenodo record and the associated article once its DOI is released.