Utilizing Optimal Transport Theory to Model Peptide Conformational Distributions and Address the Levinthal Problem [HBNI Th275]

Abstract

Peptide conformation studies are essential due to their role in biological functions like cell signaling and drug design, as well as their importance in protein structure prediction. Peptides form secondary structures such as alpha helices and beta hairpins, which can serve as building blocks for predicting three-dimensional protein structures. However, peptides exhibit structural flexibility, adopting a range of conformations, with only specific low-energy conformations being bioactive for particular functions. Constructing conformational distributions for longer peptides is challenging due to limited data from experimental sources like the Protein Data Bank (PDB), which mainly provides information for shorter peptides like dipeptides and tripeptides. In this thesis, we address this challenge by using optimal transport techniques to construct conformational distributions for longer peptides. Starting with dipeptide distributions, we develop a method to generate tetrapeptide conformational distributions by minimizing the expectation value of interaction energy functions. Applying this approach to tetrapeptides composed of alanine and glycine reveals preferences for right-handed alpha helices in alanine-rich sequences (e.g., AAAA, AAAG) and beta turns in glycine-dominated ones (e.g., GGGG, GAGG). Extending this method recursively, we generate conformational probabilities for longer peptides, enabling efficient prediction of their structural behavior. This approach provides an innovative solution for exploring peptide flexibility and bioactive conformations.