Shruti Patel

Evaluation of Vector Representations of Lipid Nanoparticles in Cheminformatic Predictions of Transfection Efficiency

Started Sept. 11, 2024

Abstract or project description

Shruti will evaluate the performance of various molecular representations in predicting the transfection efficiency of lipid nanoparticles (LNPs). By leveraging cheminformatics approaches, she will assess how different descriptors—such as molecular fingerprints, graph-based embeddings, and calculated molecular descriptors—impact model accuracy. A dataset of experimentally measured LNP transfection efficiencies will be used to train and test several machine learning representations as inputs to a neural network model. Her findings will highlight the strengths and limitations of each representation, and will offer insight into how best to represent formulations using molecular-level representations.