Predictive modeling and ranking of physicochemicals in Aegle marmelos using topological indices and multi-criteria...

Aegle marmelos (Bael Patra) contains diverse bioactive phytochemicals with significant therapeutic potential. However, it is observed that systematic computational methods for their evaluation remain limited. In this study, an integrated framework has been devised which combines quantitative structure-property relationship(QSPR) modelling and multi-criteria decision-making (MCDM) techniques to analyse and rank 15 key phytochemicals. A set of six physicochemical properties, such as molecular weight (MW), topological polar surface area (TPSA), hydrogen bond donor (HBD) and acceptor (HBA) counts, melting point (MP), and rotatable bond count (RBC) were modelled using 16 adriatic topological indices through pairwise multiple linear regression. The statistically robust models with R² ≥ 0.99 were retained. The frequency of the occurrence of descriptors in these models was used to derive objective weights, which were used in VIKOR,SAW and TOPSIS methods for ranking of compounds. The results demonstrate high predictive accuracy, with MW showing the strongest predictability and the inverse sum indeg (ISI) index emerging as the most influential descriptor. The three MCDM methods showed strong agreement, consistently identifying riboflavin, ellagic acid, and kaempferol as the top-ranked compounds. The unified approach, linking regression-based descriptor importance with decision making techniques, provides a robust and scalable approach for compound prioritization. The study possesses potential applications in virtual screening and early-stage drug discovery of plant derived molecules.