Mechanistic Studies of CRISPR-Cas9 Genome Editing Using Molecular Simulations and AI-Driven Approaches

About This Grant

Abstract Gene editing technologies based on CRISPR-Cas9 are transforming the landscape of molecular medicine by enabling targeted manipulation of nucleic acids. However, critical challenges remain in understanding the fundamental mechanisms that govern their activity, fidelity, and dynamics. The PI’s laboratory has been at the forefront of computational gene editing research, initiating the first molecular simulation of CRISPR-Cas9. Through the integration of advanced computational methods – including free energy simulations, quantum mechanical methods, cryo-EM data processing, and graph theory-based models – the lab has uncovered critical insights into conformational dynamics, catalytic mechanisms, and target specificity. These efforts have also led to the design of novel variants with improved specificity in RNA-targeting CRISPR-Cas systems. Over the next five years, the PI research program will advance the field by integrating physics-based molecular simulations with emerging methods based on Artificial Intelligence (AI). This includes deep learning-based enhanced sampling, Graph Neural Networks (GNNs) to build predictive models based on graph-structured data, and novel causality inference approaches for simulation analysis. Through these methodologies, the lab aims to provide new mechanistic characterization of the conformational rearrangements of CRISPR-Cas9, determine the structure-function relationships in prime editors, and decode how engineered variants achieve enhanced specificity. These innovations will support the broader goal of enhancing the precision and efficiency of gene editing technologies for therapeutic use. The lab’s vision is to fully integrate molecular simulations with deep learning approaches to address emerging questions in the biophysics of gene editing systems, providing novel insights that could lead to a paradigm shift in understanding their function and optimizing their functionality. By leveraging state-of-the-art computational methods integrated with AI – methods never applied before to CRISPR- Cas systems – we will pioneer new approaches to advancing genome editing technologies. This program reflects a long-term vision to establish a computational framework that not only explains existing CRISPR behaviors but also guides the design of future gene editors, in close collaboration with leading experimentalists.