Research Projects

Many key phenotypes are dynamic and invisible to endpoint readouts. We developed RainBar for pooled live-cell imaging with single-cell resolution (bioRxiv, 2025). Our CellPool (bioRxiv, 2025)efforts extend this into scalable pooled imaging screens: robust decoding, phenotyping pipelines, and stable designs at high library complexity.

To reduce the cost and logistics of large pooled screens, we’re developing multiplexed perturbation strategies that preserve interpretability. Recent results show multiplexing can enable scalable pooled screens (bioRxiv, 2025). We’re building PolySpecific CRISPR and integrating multiplexing with durable gene silencing (in progress, 2026).

We’re building high-throughput assays that connect peptide–MHC presentation to T cell receptor recognition and functional readouts. The goal is to make immune specificity measurable and engineerable, and to generate pMHC–TCR datasets that enable predictive models that generalize (in preparation, 2026).

ExoRelay engineers immune cells to package RNA barcodes into extracellular vesicles that can be captured and sequenced from accessible biofluids, enabling longitudinal readouts of immune activity from small samples (in preparation, 2026).

Cell Relay maps direct cell–cell contact via contact-dependent RNA transfer mechanisms, enabling reconstruction of interacting pairs by sequencing. It’s designed for immune recognition, where stable contact encodes specificity (in preparation, 2026).

We’re developing approaches to quantify commensal microbes and bacteriophage in vivo across time, tissues, and perturbations, and to relate these dynamics to host state transitions and immune outcomes. These studies are in progress—please contact us if you’re interested in this area.

Early work established RNAi in primary T lymphocytes (J Immunol, 2002) and one of the first descriptions of shRNA technologies (RNA, 2002). These studies led to the identification of the first mammalian endogenous siRNA (RNA, 2002). We also built resources for studying microRNA in vivo (Cell Reports, 2012) and participated in the first microRNA mouse knockout (Cell, 2007).

We helped define how chromatin state organizes the human genome into interpretable functional elements, including participation in the integrative analysis of 111 reference human epigenomes (Nature, 2015). We also contributed a lincRNA survey supporting regulated promoter logic and chromatin-linked control (PLoS Genetics, 2013).

We conducted genome-scale screening with colleagues to identify endogenous barriers resisting iPSC generation (Cell, 2014) and characterization of CHD1 that maintains an open, pluripotency-compatible chromatin state (Nature, 2009). We also contributed to work on transition-like potency states (Science Advances, 2025) and chromatin accessibility dynamics (Stem Cells, 2023).

We co-developed early mammalian genetic interaction mapping in pooled formats (Cell, 2013) and advanced combinatorial perturbation approaches that map directional dependencies at scale (Nature Biotechnology, 2018). In cancer, we helped define the drug-tolerant persister state and discovered vulnerability to ferroptotic death through GPX4 inhibition (Nature, 2017).

We developed a CRISPR interference (CRISPRi) mouse model based on dCas9-KRAB for stable, tissue-specific gene silencing. We applied this system to neurons (Neuron, 2019) and metabolic tissues (Nature, 2019), reinforcing a lab value: build broadly useful resources through team science.

Our early work spans diverse areas-from RNA technologies to regulatory genomics resources to genetic interaction mapping and in vivo CRISPRi—with a common thread has been building quantitative platforms that scale beyond our own lab. We value open collaboration, resource sharing, and methods that make hard biology tractable for many groups.