Seminars

Lipid nanoparticles (LNPs) are a leading nonviral nucleic acid delivery technology. Optimizing the constituent lipids of LNPs is critical for improving delivery efficacy, yet their underlying structure-function relationship is incompletely understood. Here we develop data-driven approaches to better understand and improve mRNA LNPs. First, we introduce Lipid Nanoparticle Database to advance deep learning and simulation approaches. Second, we develop a self-driving lab for the directed evolution of LNPs across three distinct delivery challenges.

Optical pooled screening (OPS) is a technique for exploring genotype-to-phenotype relationships, where cells are genetically perturbed and imaged to study morphological changes. While fluorescence microscopy is often used, non-perturbative ‘label-free’ imaging presents several advantages. A promising approach is quantitative phase imaging, which provides contrast based on a cell’s refractive index map. We conducted an OPS screen, pairing label-free methods with 21-plex fluorescence imaging to explore their potential in morphological profiling.