Developing Superior Amoxicillin Delivery Systems: AI-Driven Optimization of LNPs for H. pylori Treatment

Abstract The development of nano delivery systems, particularly lipid nanoparticles (LNP), for both hydrophobic and hydrophilic drugs has seen significant advancements in recent years. Fine tuning LNP formulations is crucial due to the impact of various parameters on their quality of efficacy. The study investigated the influence of formulation variables on amoxicillin-loaded LNPs designed for anti-Helicobacter pylori activity. Size, polydispersity index (PDI), Zeta potential and entrapment efficiency were evaluated across diverse formulations. The impact of particle size on drug release and encapsulation was explored. Artificial intelligence AI based design of experiments generated formulations to minimize the particle size, PDI and Zeta potential while maximizing the EE, accounting for factor interactions. Additionally, the user friendliness of QbD (Quality by Design), Machine Learning (ML), and DOE were compared. Methods and results: A Box-Behnken design with 27 formulations was chosen for amoxicillin (amox) LNP optimization. Particle size distribution, zetapotential, PDI, and entrapment efficiency were measured for each formulation. LNP ranged in size from 200–600 nm, zeta potential ranged from − 5 - -40 mV, PDI from 0.1- 1 and EE from 5-100%. Characterization included DLS, FESEM, FTIR and SEM. Obtained results were statistically analysed. Discussion: This study demonstrates the potential of AI- driven DOE for optimizing LNP formulations. We explained effect of different parameters lipid concentration, surfactant concentration, sonication time and sonication speed on nanoparticles and derived formula for further prediction. The identified formulations exhibited desired antibiotic efficiency with minimum chemical usage, suggesting the effectiveness of this approach. Further research explored it as a drug with more bioavailability, stability and cheap alternative over traditional drugs in market with more side effects and less bioavailability.