A Novel Approach for Lipophilicity-Based Genetic Delivery Formulation of Cystic Fibrosis Therapeutics via LNP-VACCO: Lipid Nanoparticle Variational Autoencoder-Guided Combinatorial-Chemistry Optimization

EasyChair Preprint no. 13145

Abstract

Cystic Fibrosis (CF), characterized by its profound impact on respiratory and digestive functions, arises due to genetic mutations in the CFTR gene on chromosome-7. Despite progress in medical science, treatments often offer incomplete restoration of chloride function and are burdened by complications and side effects, highlighting an unmet medical need. The emergence of gene editing technologies, particularly involving chemically-modified-mRNA, shows promise in addressing these underlying genetic mutations. Concurrently, Lipid Nanoparticles (LNPs) have revolutionized the pharmaceutical industry, with mRNA-based therapies being at the forefront of innovation. However, the formulation of LNPs presents challenges concerning stability and biocompatibility. This research introduces LNP-VACCO, a novel approach integrating Variational Autoencoders and Combinatorial-Chemistry. By leveraging principles of lipophilicity and Simplified-Molecular-Input Line-Entry-System (SMILES) strings, LNP-VACCO autonomously navigates the landscape of LNP compositions, offering efficient explorations of potential formulations. The methodology involves three-step deep-learning processes, wherein the model iteratively refines lipid constituent compositions to optimize LNP performance. In-vitro validation experiments, involving synthesis and transfection of lipids into HeLa cells to simulate CF conditions, demonstrated promising results regarding encapsulation efficiency and cell viability. LNP-VACCO, a significant leap forward in enhancing the efficacy of nanoparticle-based drug delivery systems, offers hope for effective treatments for CF and other genetic disorders.

Keyphrases: Combinatorial Chemistry, Cystic Fibrosis, deep learning, drug discovery, Genetic Editing, In-Vitro Validation, lipid nanoparticles, Unsupervised Machine Learning

@Booklet{EasyChair:13145,
  author = {Nidhi Yadalam},
  title = {A Novel Approach for Lipophilicity-Based Genetic Delivery Formulation of Cystic Fibrosis Therapeutics via LNP-VACCO: Lipid Nanoparticle Variational Autoencoder-Guided Combinatorial-Chemistry Optimization},
  howpublished = {EasyChair Preprint no. 13145},

  year = {EasyChair, 2024}}