Chen YuLan, Rauf Abdul, Shafique Aqsa, Tchier Fairouz, Aslam Adnan, Tola Keneni Abera
Department of Neurology, Nanjing Pukou Hospital of TCM, Nanjing, China.
Department of Mathematics, Air University Multan Campus, Multan, Pakistan.
Sci Rep. 2025 May 3;15(1):15527. doi: 10.1038/s41598-025-99583-8.
Parkinson's disease is a progressive neurological disorder characterized by the degeneration of the nervous system, leading to impaired motor and non-motor functions. Early symptoms include tremors, rigidity, and bradykinesia, with progressive deterioration over time. This study employs a multi-criteria decision-making approach, integrating Fuzzy TOPSIS and Quantitative Structure-Property Relationship (QSPR) analysis, to evaluate and rank 17 Parkinson's disease medications based on their physicochemical properties. Molecular structures were encoded as adjacency matrices using MATLAB 2017, and six Sombor index variants-computed via a custom Maple 2020 algorithm-served as topological descriptors for QSPR modeling. Eight critical physicochemical properties were analyzed: polarizability (P), boiling point (BP), surface tension (ST), polar surface area (PSA), flash point (FP), molar refractivity (MR), enthalpy of vaporization (EV), and molar volume (MV). The Fuzzy TOPSIS ranking revealed bromocriptine as the top-performing drug for boiling point (BP), while comparative rankings across all properties are tabulated for clinical reference. Validation metrics, including coefficient of determination, mean squared error , and mean absolute error, confirmed model robustness. Notably, surface tension (ST) and polar surface area (PSA) showed weaker correlations (R < 0.5, p > 0.05), highlighting limitations in their predictability via Sombor indices. This work demonstrates the utility of combining chemical graph theory, QSPR modeling, and Fuzzy TOPSIS for rational drug evaluation in neurodegenerative disorders. The methodology offers a framework for prioritizing therapeutics based on physicochemical profiles, with implications for optimizing Parkinson's disease management.
帕金森病是一种进行性神经疾病，其特征是神经系统退化，导致运动和非运动功能受损。早期症状包括震颤、僵硬和运动迟缓，随着时间的推移会逐渐恶化。本研究采用多标准决策方法，将模糊理想解排序法（Fuzzy TOPSIS）与定量构效关系（QSPR）分析相结合，根据17种帕金森病药物的物理化学性质对其进行评估和排名。使用MATLAB 2017将分子结构编码为邻接矩阵，并通过自定义的Maple 2020算法计算出六种索博尔指数变体，作为QSPR建模的拓扑描述符。分析了八个关键的物理化学性质：极化率（P）、沸点（BP）、表面张力（ST）、极性表面积（PSA）、闪点（FP）、摩尔折射率（MR）、汽化焓（EV）和摩尔体积（MV）。模糊理想解排序法排名显示，溴隐亭是沸点（BP）方面表现最佳的药物，同时列出了所有性质的比较排名以供临床参考。包括决定系数、均方误差和平均绝对误差在内的验证指标证实了模型的稳健性。值得注意的是，表面张力（ST）和极性表面积（PSA）显示出较弱的相关性（R < 0.5，p > 0.05），突出了通过索博尔指数对其进行预测的局限性。这项工作证明了将化学图论、QSPR建模和模糊理想解排序法相结合用于神经退行性疾病合理药物评估的实用性。该方法提供了一个基于物理化学特征对治疗方法进行优先排序的框架，对优化帕金森病管理具有重要意义。
