Our lab has been exploring the use of protein language models coupled with high-throughput protein variant screens to rapidly generate optimized protein variants. We are interested in applying these tools towards the amelioration of proteins misfolding and the optimization of gene editing tools.

We have developed two novel technologies for rapidly generating pools of thousands of CRISPR variants and assaying their activities in a single study. Using these methods we aim to improve the efficacy of existing CRISPR tools along with endowing CRISPR tools with new activities.

High rates of failed implantation in human embryos represents one of the greatest obstacles in our ability to treat infertility. Significant improvements require significant advances in our understanding of the molecular mechanisms required for successful implantation and ongoing development. The goal of this funded project is to uncover the network of regulatory factors and cell-cell signals that control cellular differentiation, developmental progression and successful implantation during early human embryo development using both human embryo and stem cell-based in vitro models. These analyses will provide the first “ground truth” atlas for human embryos of high implantation potential—and for stem cells developed to model differentiation within the embryo--and essential first steps toward development of an in vitro model for human embryo implantation.

mRNA polyadenosine (polyA) tail lengths play a unique and critical role in controlling gene expression in the developmental transition from oocyte to embryos from worms to humans. We have recently generated a large comprehensive profile of polyA tail lengths across the mouse oocyte-to-embryo transition with Nanopore long read sequencing, capturing tail length regulation with isoform-specific resolution (including 3'UTR length, splice isoform, polyadenylation site choice, etc.). We found dynamic changes in polyA tail length across this transition and that these changes in tail length control mRNA translation and stability. But what molecular mechanisms orchestrate these changes in polyA tail length?  In this project, we will apply machine learning approaches to ask which mRNA features (number and position of specific RNA binding protein motifs, mRNA length, mRNA abundance, polyadenylation site choice, 3'UTR length, etc.) are most predictive of (1) polyA tail length at each developmental stage and (2) change in tail length across consecutive developmental stage transitions.  These analyses provide an exciting opportunity to address questions decades-old questions as to the mechanisms driving the earliest stages of mammalian development.