1. Using data from a high-throughput experimental assay, we have developed a computational algorithm that predicts the genotypic outcomes of template-free CRISPR/Cas9 genome editing. This modeling has uncovered that CRISPR/Cas9 outcomes are highly predictable and often skewed toward a single outcome, including at human pathogenic indel mutation sites. We are now working to improve the precision of CRISPR/Cas9 editing and to employ it in novel genomic screening platforms to identify gene therapy strategies.

2. We are developing predictive computational models of transcription factor binding. To facilitate this goal, we have developed a technique to insert libraries of 10,000s of DNA sequences into a fixed locus in the genome using CRISPR homologous recombination. Then we probe how each sequence is bound by a transcription factor of interest and how that translates to epigenetic features such as chromatin accessibility.

3. It is currently not possible to reprogram patient cells to a majority of the cell types of interest for disease modeling and regenerative medicine, and we have built a platform to use combinatorial transcription factor delivery and single cell RNA-seq to identify which transcription factor combinations can reprogram cells to cell fates of therapeutic interest. There are exciting implications in predictive understanding of cell fate and in drug/gene screening in reprogrammed cell types.