Biotransport dynamics of MXene nanoparticles in Casson micropolar flow: From Lie symmetry analysis to stacking ensemble learning

Targeted drug delivery within a transport medium of Casson micro-polar nanofluid flowing over endothelial cells guarantees effective therapy with minimal side effects, yet intricate nano-bio interactions, reduced delivery performance, and safety concerns often limit its efficiency. In this study, stable and biocompatible MXene (Ti₃C₂) nanoparticles derived from MAX-phase carbide/nitride materials are employed as drug carriers, guided by an inclined magnetic field, Newtonian heating, microorganisms, and relaxation controls via Cattaneo-Christov heat and mass flux theory to enable precise, site-specific drug release. Hemodynamic forces induce vessel wall expansion and endothelial stretch, which are represented mathematically as a stretching sheet, enabling the capture of dynamic biotransport interactions across the blood milieu and the endothelial wall. The governing partial differential equations are reduced to ordinary differential equations via Lie symmetry analysis and solved numerically using the Boundary Value Problem 5th-order Collocation solver. Results show that microrotation and magnetic strength enhance particle velocity and drug penetration, while higher chemical reaction rates decrease concentration. Increased solutal relaxation delays diffusion, thereby improving absorption. To complement the theoretical analysis, a stacked ensemble combining linear regression, random forest, and gradient boosting achieved superior predictive accuracy (R²=0.9793,RMSE=0.0091 for skin friction; R²=0.9980,RMSE=0.0020 for mass transfer). Shapley Additive exPlanations (SHAP) sensitivity analysis revealed that Casson, solutal relaxation time, and chemical reaction parameters were the most contributing features to the skin friction coefficient. This combined analytical-computational framework provides new insights into nanoparticle-assisted, site-specific drug delivery under complex hemodynamic conditions.