Computational Discovery of Antithrombotic Ferulate Analogs Through QSPR Modeling, Pharmacokinetic Prediction, Molecular Docking, and Dynamic Stability Assessment
نویسندگان
1 Doctoral Progamme of Pharmaceutical Sciences, Faculty of Pharmacy, Universitas Airlangga, Surabaya, Indonesia
2 Department of Pharmaceutical Chemistry, Faculty of Pharmacy, Universitas Hang Tuah, Surabaya, Indonesia
3 Department of Pharmaceutical Sciences, Faculty of Pharmacy, Universitas Airlangga, Surabaya, Indonesia
4 Drug Development Research Group, Faculty of Pharmacy, Universitas Airlangga, Surabaya, Indonesia
5 5Inter-University Center of Excellence (IUCoE) of Health Autonomy-Drug Discovery, Universitas Airlangga, Surabaya, Indonesia
6 Department of Pharmaceutical Sciences, Faculty of Pharmacy, Universitas Airlangga, Surabaya, Indonesia
doi
10.48309/chemm.2026.563470.2055چکیده
Stroke remains a major global health burden, motivating the discovery of safer and more effective antithrombotic candidates. Antithrombotic therapy such as aspirin is widely used; yet, its clinical benefit is constrained by gastrointestinal bleeding risk and interindividual variability, highlighting the need for safer and more effective candidates. Ferulic acid exhibits antithrombotic potential, but its potency is generally lower than that of standard therapy. In this study, an integrated in silico workflow Topliss-guided design, drug-likeness/ADMET screening, network pharmacology, structure-based docking and molecular dynamics (MD), as well as QSPR analysis of Log CLtot were implemented to computationally prioritize a ferulate-derived lead for stroke-related thrombosis. The SAR space was intentionally focused on selective modification of the phenolic –OH via o-benzoylation, yielding 33 synthetically accessible analogs while preserving the cinnamate core. Network pharmacology (SEA, SwissTargetPrediction, BindingDB, DisGeNET, OMIM, STRING, Metascape, and KEGG) prioritized the thromboxane A₂ receptor (TXA₂R/TBXA2R) as a key node based on high network centrality and mechanistic relevance to platelet activation. Molecular docking in MOE, validated by redocking (RMSD = 0.7404 Å), highlighted 4-(3-trifluoromethylbenzoyloxy)-3-methoxycinnamic acid as the top-ranked candidate (S-score = −10.0556 kcal/mol). A 100-ns MD simulation supported stable binding in the TXA₂R pocket, and post-MD MM-GBSA estimated a favorable binding free energy (ΔGbind = −57.65 kcal/mol). QSPR analysis suggested that Log CLtot is associated with aqueous solubility (LogS), with additional contributions from electronic (Etot) and steric (CMR) descriptors; predictive validation and robustness testing are addressed in the revised manuscript to avoid overinterpretation. Overall, this study computationally prioritized a ferulate-based antithrombotic lead to support synthesis and experimental validation.