Computational Approaches to Understanding the Structural Features of Small Molecules of Biological Interest

Abstract

Pancreatic ductal adenocarcinoma (PDAC) stands out as one of the most aggressive and lethal forms of cancer, characterized by a dismal prognosis largely due to late-stage diagnosis and the inadequate effectiveness of current treatment options. Recognizing the urgent need for innovative therapies, this thesis adopts a dual approach, exploring both direct and indirect inhibition mechanisms to tackle the challenges associated with PDAC treatment. By integrating computer-aided drug design with the chemical diversity of natural products, novel small-molecule inhibitors with potential efficacy against PDAC were identified. Divided into two main parts, the study begins with an exploration of curcumin derivatives as potential inhibitors of PI3Kα. Through virtual screening that employed pharmacophore modeling and molecular docking, two promising compounds were identified: CID154728220 and CID156189304. Both compounds exhibited favorable pharmacokinetic profiles, and subsequent MD simulations confirmed their structural stability, making them strong candidates for further preclinical evaluation. The second part utilized a fragment-based drug discovery approach to directly target KRAS G12D. This strategy led to the discovery of two novel compounds, Hit1 and Hit2, which demonstrated higher binding affinity compared to the reference inhibitor MRTX 1133, as validated by IFD and MM-GBSA analysis. Notably, Hit2 exhibited the most favorable balance of pharmacokinetic properties, safety, and drug-likeness, positioning it as a promising candidate for further development. This thesis represents a significant advancement in cancer research, providing a foundation for the development of novel therapeutic strategies in the treatment of pancreatic cancer