Publications
* co-first author ^ co-corresponding author
Multiomics and deep learning dissect regulatory syntax in human development
My main post doc work using deep learning on a new fetal single-cell atlas to learn about the organization of transcription factor binding sites in human development, and where they're perturbed in complex traits. Highlighted in Nature Genetics.
Sensitive, direct detection of non-coding off-target base editor unwinding and editing in primary cells
We present a new assay, beCasKAS, to sensitively detect off-target edits from CRISPR base editors, and show how non-coding off-target genome edits can be triaged for epigenetic dysregulation in silico using deep learning.
A unified lexicon of predictive DNA sequence motifs from ENCODE transcription factor binding and chromatin accessibility assays
FOXR2 targets LHX6+/DLX+ neural lineages to drive CNS neuroblastoma
We analyzed single-cell profiles of the developing brain together with molecular profiles from patient samples to nominate cell-of-origin, which enabled rationally guided mouse modeling of FOXR2-activated CNS neuroblastoma, a rare pediatric brain tumor subtype. See a commentary on our approach and findings here.
TULIPs decorate the three-dimensional genome of PFA ependymoma
K27M in canonical and noncanonical H3 variants occurs in distinct oligodendroglial cell lineages in brain midline gliomas
Using single-cell and bulk transcriptomics and epigenomics, we provide molecular evidence that deadly childhood gliomas with K27M mutations in different Histone 3 variants arise from distinct anatomical domains in the developing brain. See a News & Views on ours and complementary work here.
Histone H3.3G34-Mutant Interneuron Progenitors Co-opt PDGFRA for Gliomagenesis
We show that despite being histologically considered gliomas, a subtype of Histone 3 mutant childhood brain tumors likely arise in the interneuron lineage in the developing telencephalon; and that these tumors co-opt the PDGFRA oncogene to drive tumorigenesis
Stalled developmental programs at the root of pediatric brain tumors
We generate a single-cell RNAseq atlas of developing brain regions where pediatric brain tumors occur, and use a data-driven approach to systematically nominate cell-of-origin. This work was selected as one of the top 10 discoveries in Quebec for 2020!
Healthcare for All!: Access to healthcare for migrants with precarious status in the Parc-Extension neighbourhood of Montreal, Quebec