Merci Ismael 😉

We are also grateful to the @broadinstitute.org and the @GRO_Broad for their support, and to our editor and reviewers at @science.org! (end)

I want to sincerely thank all our amazing co-authors and, in particular, my wonderful co-first author Jorge Martin-Rufino (@jmartinrufino), as well as my mentors @bloodgenes.bsky.social and @eric_lander. (12/n)

Our findings reveal that heritable variation is predominantly enriched in specific gene regulatory networks controlled by master TFs, offering novel insights into the genetic architecture of complex traits. Protocols and much more can be found in our manuscript! (11/n)

Most strikingly, we discovered that TF-sensitive regions comprise <0.3% of the genome but are enriched ~100-fold for SNPs associated with certain blood-cell phenotypes, which is remarkably higher than that of any other chromatin elements active throughout hematopoiesis. (10/n)

…such as a GATA1-dependent regulatory network comprising the CPEB4 gene and an enhancer harboring variants with multiple associations to blood cell traits, exemplifying how Perturb-multiome can help systematically pinpoint associations between phenotypes and GWAS variants. (9/n)

For example, we faithfully identified various well-characterized gene regulatory mechanisms in hematopoiesis, such as the BCL11A-mediated fetal-to-adult hemoglobin switch. Importantly, it also uncovered thousands of previously uncharacterized regulatory interactions… (8/n)

As such, Perturb-multiome allowed us to reconstruct transcription factor-dependent gene regulatory networks throughout hematopoietic differentiation, capturing the complexity of the differentiation continuum and enabling a detailed understanding of regulatory mechanisms. (7/n)

Perturb-multiome is highly effective in disease-relevant primary cells, such as hematopoietic stem and progenitor cells (which give rise to all blood cells) by overcoming challenges in perturbation efficiency and identification. (6/n)

With Perturb-multiome we can now generate high-resolution analyses of gene regulatory networks in single cells, which enables genome-wide variant-to-function inferences. (5/n)

To achieve this, we developed Perturb-multiome, which recovers from each cell: (1) the identity of the genetic perturbation (sgRNA), (2) the effect on chromatin accessibility at cis-regulatory elements (scATAC-seq), and (3) the effect on expression of genes (scRNA-seq). (4/n)

Rather than recreating each variant, which is hard to scale, given the enrichment of noncoding causal variants in regulatory regions bound by transcription factors (TFs), we reasoned that systematic perturbation of TF networks in single cells could provide new insights (3/n)

While GWAS have been successful in identifying a large number of genetic variants associated with complex traits, biological insights gained have been limited. This is largely due to the lack of systematic methods for linking these variants to their functional impacts. (2/n)