Structure-Activity Relationship (SAR) as a Weapon Against Chemoresistance in Pancreatic Cancer

Pancreatic ductal adenocarcinoma (PDAC) is one of the most morbid malignancies with drug resistance to chemotherapy as an alarming challenge. Structure-Activity relationship (SAR) approaches have emerged as useful computational tools to predict and design novel chemotherapeutic agents that would overcome the resistance mechanisms. This review compiles recent advances in SAR based methodologies over the last five years, applied to research in PDAC’s drug resistance and evaluates computational approaches, experimental validations and translational potential. Recent SAR studies have shown progress in three main avenues: ligand-based methods using QSAR, 3D-QSAR and pharmacophore modelling; structure-based approaches integrating molecular docking with binding site analysis; and machine learning (ML) approaches utilizing artificial intelligence (AI); to combat the ever-increasing fatality of pancreatic cancer. However, limitations and drawbacks exist in these methods which include issues in validation with small dataset, risks of overfitting in ML models, restricted applicability domains and significant gaps in translation between computational predictions and clinical outcomes. More so, most SAR studies have not yet explored the pharmacokinetic properties, effects of tumor microenvironment (TME) and stem cell targeting approaches. The field of SAR is nevertheless advancing towards integration of multi-omics data, improvement of experimental validation by using pre-clinical models; combination of SAR with nanomedicine delivery systems and development of AI-powered personalized therapeutic strategies which target specific resistance mechanisms.