-
What are all those peaks doing?
Swap-seq uses pooled twin prime editing to tile precise, barcoded deletions across a locus — finding enhancers, CTCF sites, and a distal HDAC-dependent silencer that CRISPRi and AI sequence models miss.
-
scE2G: linking enhancers to their target genes from single-cell data
A model that predicts which enhancers regulate which genes using single-cell ATAC or multiome data — outperforming existing single-cell methods, and opening the way to enhancer maps across the thousands of cell types in the human body.
-
Building an encyclopedia of which enhancers regulate which genes
Six years, 1,458 biosamples and 92 million predicted enhancer–gene links — how the ENCODE-rE2G model was benchmarked into existence, and what it taught us about the rules of enhancer–gene regulation.
-
Does every enhancer work with every promoter?
An enhancer × promoter MPRA, rebuilt to remove assay confounders, shows that promoters differ more than 100-fold in their intrinsic responsiveness to enhancers — reconciling a long-standing contradiction in the field.
-
Why guinea pigs grow natural bypass arteries — and how we might too
An in vivo Perturb-seq screen, guided by the differences between species, uncovers the pathways that build protective collateral arteries.
-
The hidden catch in CRISPR-Cas13d screens for functional circular RNAs
Junction-targeting Cas13d guides often knock down the linear mRNA too — so many 'functional circRNA' hits may be off-target artifacts. A cautionary re-analysis, now in Nucleic Acids Research.
-
What we learn by CRISPR'ing a random set of regulatory elements
An unbiased CRISPRi survey of ~1,000 randomly chosen accessible sites — and a new toolkit (DC-TAP-seq) for measuring enhancer effects on gene expression.
-
The E2G Portal: enhancer–gene predictions across 1,600+ cell types
A major milestone in mapping enhancers and interpreting variants — browse our enhancer-gene predictions across more than 1,600 cell types and tissues at e2g.stanford.edu.
-
Rewriting regulatory DNA to dissect and reprogram gene expression
Variant-EFFECTS uses high-throughput prime editing, flow sorting, and sequencing to precisely measure how noncoding variants change gene expression — now in Cell.
-
What cell types drive congenital heart defects?
A single-cell multiomic atlas of 734,000 cells from 41 fetal hearts — and what it reveals about the cell types and noncoding variants behind congenital heart disease.
-
How do 41 scattered GWAS loci turn out to be one pathway?
Variant-to-Gene-to-Program (V2G2P) combines an enhancer–gene map with Perturb-seq to trace coronary artery disease variants to the genes and pathways they act through — and finds them converging on a single endothelial complex.
-
Can a formula this simple predict which enhancers regulate which genes?
Thousands of CRISPR perturbations showed that every existing model for enhancer–gene regulation performed poorly — and that one multiplying enhancer activity by 3D contact performed remarkably well.
-
When a gene's promoter turns out to regulate its neighbour
Heterozygous knockouts across 18 loci showed that half of the gene promoters tested regulate a neighbouring gene — acting as enhancers or as competitive repressors — while 12 lncRNAs tested had no detectable cis function.