Per- and polyfluoroalkyl substances (PFAS) are environmentally persistent organofluorines linked to cancer, organ dysfunction, and other health problems. This study used quantitative structure–property relationship (QSPR) and quantitative structure–activity relationship (QSAR) modeling to examine the binding of PFAS to estrogen receptor alpha (ERα) and beta (ERβ). Molecular docking of 14,591 PFAS compounds was performed, and docking scores were used as a measure of receptor affinity. QSPR models were built for two datasets: the ERα and ERβ top binders (TBs), and a set of commonly exposed (CE) PFAS. These models quantified how chemical descriptors influence binding affinity. Across the models, higher density and electrophilicity indicated positive correlations with affinity, while surface tension indicated negative correlations. Electrostatic descriptors, including HOMO energy and positive Fukui index (F⁺ max), were part of the models but showed inconsistent trends. The CE QSPR models displayed correlations that conflicted with those of the TB models. Following QSPR analysis, 66 QSAR models were developed using a mix of top binders and experimental data. These models achieved strong performance, with R² values averaging 0.95 for training sets and 0.78 for test sets, that indicated reliable predictive ability. To improve generalizability, large-set QSAR models were created for each receptor. After outlier removal, these models reached R² values of 0.68–0.71, which supports their use in screening structurally diverse PFAS. Overall, QSPR and QSAR analyses reveal key chemical features that influence PFAS–ER binding. This predictive approach provides a scalable framework to assess the binding interactions of structurally diverse PFAS to ERs and other nuclear receptors. All the codes, data, and the GUI visualization of the results are freely available at sivaGU/QSPR-QSAR-Molecular-Visualization-Tool.

Metalloprotein docking is difficult because it requires precise preparation of the grid box, charge assignment, and metal coordination geometry. Many current graphical user interfaces (GUIs) do not support these steps well. Here, we present MBind v2026, a web-based GUI that combines AutoDock Vina, AutoGrid4 and AutoDock4 into a single workflow for standard docking and metalloprotein docking. The interface also has the optional machine learning (ML) docking program GNINA v1.0. The main validated metalloprotein workflow in MBind is based on AutoDock4Zn parameterization for zinc, while preliminary workflows for magnesium, iron, and copper are also included. MBind was benchmarked on eight zinc metalloproteins and eight nonmetalloprotein targets. Its performance was compared with ML pose prediction methods (GNINA v1.0, EquiBind v2026, TankBind v2026, and GAABind v2026), cofolding models (Boltz-2 v2026, and AlphaFold 3 v2026), ML affinity prediction tools (StructureNet v2026, GNNSeq v2026, and PLAIG v2026), and non-ML docking tools (SwissDock v2026, CB-Dock2 v2026, Webina v2026, 1-ClickDock v2026, and MolModa v1.01). Pose accuracy was measured by symmetry-aware ligand RMSD using DockRMSD v1.1 under redocking conditions with co-crystal-defined binding sites. Binding energy trends were evaluated by comparing docking scores with IC50-derived ΔG values using mean absolute error. On the zinc metalloprotein benchmark set, MBind produced a mean RMSD of 0.49 Å. On the nonmetalloprotein set, the mean RMSD was 0.62 Å. These results show that MBind can execute zinc metalloprotein and nonmetalloprotein docking workflows through a web-based interface, while the preliminary Mg, Fe, and Cu workflows require additional validation. The MBind GUI web-based platform is publicly available.

Per- and polyfluoroalkyl substances (PFAS) can bioaccumulate and may increase the risk of cancer development. Clinical studies report that cholestyramine treatment lowers serum PFAS concentrations, but the molecular energetics, interactions and underlying mechanisms of PFAS exchange with the resin remain incompletely understood. Here, we present a quantum perspective to define the mechanism of PFAS anion exchange with cholestyramine. Quantum chemistry calculations, including density functional theory (DFT), were used to evaluate chloride exchange for perfluorooctanesulfonic acid (PFOS), perfluorooctanoic acid (PFOA), perfluorohexanoic acid (PFHxA), and 6:2 fluorotelomer carboxylic acid (6:2 FTCA) at an isolated ammonium site and in an extended styrene-divinylbenzene model. Single-site ion pairing did not make exchange with chloride thermodynamically favorable. Adding the divinylbenzene backbone lowered electronic exchange energies for all four PFAS, whereas the Gibbs free-energy effect varied among compounds. Energy decomposition analysis showed that electrostatic attraction was the largest stabilizing contribution, while added backbone contacts altered dispersion, exchange correlation and Pauli contributions. For PFOA, changing the bound state continuum from water to 1-octanol produced little net change in exchange free energy. Three separately seeded 10 ns molecular dynamics trajectories of a hydrated pVBTMA₁₂ oligomer showed persistent PFOA contact with changing headgroup and tail environments, although global structural convergence was not established. Together, these results support a molecular mechanism in which electrostatic ion pairing and PFAS backbone contacts jointly influence anion exchange favorability in simplified cholestyramine models. This mechanistic framework provides testable hypotheses for PFAS sequestration and identifies aromatic contact area and ammonium site geometry as design variables for future resin optimization. These findings connect quantum level interaction energetics with experimentally testable strategies for developing improved PFAS sequestrants. Computational data and code are archived in Zenodo, and a web-based graphical user interface tool is available online.

Background/Objectives: Blood–brain barrier (BBB) permeability prediction is important for central nervous system drug discovery. Many models rely mainly on chemical structure, omit transport information, and may report optimistic performance when related compounds occur across training and test sets without adequate leakage control. We developed MechBBB, a two-stage, transport-informed machine-learning framework that adds learned efflux, influx, and passive-permeability scores to conventional molecular features and evaluates their added value under explicit leakage control. Methods: In Stage 1, three LightGBM models were trained on separate efflux, influx, and passive-permeability datasets to generate transport-related scores. BBBP compounds were excluded from Stage 1 training by InChIKey matching. In Stage 2, a LightGBM classifier was trained on BBBP using a Murcko scaffold split with the three Stage 1 scores, ten physicochemical descriptors, and 2048-bit ECFP4 fingerprints. Performance was compared with a descriptor-plus-fingerprint baseline without transport scores. External evaluation used a strict B3DB set (n = 4080) with no InChIKey or nonempty scaffold overlap with BBBP and was performed without retraining, recalibration, or threshold adjustment. MechBBB was also compared with official SwissADME BOILED-Egg predictions on a chemically matched external subset. Results: On the BBBP scaffold test set, MechBBB achieved an AUROC of 0.932, an AUPRC of 0.983, an MCC of 0.737, and a balanced accuracy of 0.866. The AUROC improvement over the descriptor-plus-fingerprint baseline was small but statistically significant in the paired test (ΔAUROC = 0.0096; p = 0.0425). Scaffold-grouped cross-validation did not show a consistent ranking advantage from adding the Stage 1 scores. On strict B3DB, MechBBB achieved an AUROC of 0.894 and an AUPRC of 0.893, with no significant ranking advantage over the baseline. On the matched B3DB subset (n = 4043), MechBBB achieved higher accuracy (0.814 vs. 0.682), balanced accuracy (0.806 vs. 0.693), sensitivity (0.910 vs. 0.541), and MCC (0.631 vs. 0.401) than SwissADME BOILED-Egg, whereas BOILED-Egg had higher specificity (0.845 vs. 0.703). SHAP analysis identified TPSA, NumHDonors, and p_pampa among the leading contributors. Conclusions: MechBBB provides a leakage controlled and externally tested framework that combines competitive BBB permeability prediction with per-compound transport-related information. The Stage 1 scores produced a modest internal improvement but did not provide a consistent external ranking advantage, indicating that their main added value is transport-related context rather than a universal increase in predictive accuracy. MechBBB can be used to prioritize compounds for CNS drug discovery before experimental BBB testing and identify compounds that may warrant follow-up studies of passive permeability or active transport. The MechBBB web tool is freely available.