Research Roundup: October 18, 2019

The case for diverse genetic data, finding there's more to GIST than mutations, pooled perturbation screens go optical, and more.

Erik Jacobs
Credit: Erik Jacobs

Welcome to the October 18, 2019 installment of Research Roundup, a recurring snapshot of recent studies published by scientists at the Broad Institute and their collaborators.

Better biology through greater genetic diversity

The low level of ancestral diversity in GWAS datasets could limit what scientists can learn about disease biology and contribute to global health disparities. In a Primer article in Cell, Hailiang Huang of the Stanley Center for Psychiatric Research and his colleagues outline methodological pitfalls and recommendations for those studying data from more diverse populations. Such considerations include genotyping technologies, data quality-control criteria, statistical methods, and more. The authors also provide a catalog of sources of diverse whole-genome sequence data, and describe areas where analytic methods need improvement. Read more in this press release from Virginia Commonwealth University.

Getting the GIST of it

Most gastrointestinal stromal tumors (GISTs, a common GI cancer) can be treated with targeted kinase inhibitors, but roughly 10 to 20 percent lack the requisite mutations. In Nature, William Flavahan, Yotam Drier, Sarah Johnstone, institute member and Epigenomics Program director Bradley Bernstein, and colleagues show that some of these tumors harbor epigenetic changes that break structural barriers within the genome meant to keep genes away from other genes’ enhancers (“on” switches). The molecular errors activate an oncogene called FGF4, and render the tumors vulnerable to drugs called FGFR inhibitors, either alone or together with sunitinib, a standard GIST treatment. Learn more in a press release from Massachusetts General Hospital.

Giving tumors the one-two punch

Chromatin regulation is altered in diffuse intrinsic pontine glioma (DIPG), an incurable pediatric cancer. A team including Jamie Anastas (Boston Children’s Hospital/HMS), associate members Mariella Filbin and Yang Shi, and core institute member and Cancer Program director Todd Golub used CRISPR screening to reveal that inhibiting histone deacetylases (HDAC) sensitizes cells to the demethylase LSD1. Reporting in Cancer Cell, they show that Corin, a dual inhibitor of HDACs and LSD1, potently inhibits DIPG growth in vitro and in xenografts. It causes transcriptional changes correlating with increased survival time in patients, suggesting that combination epigenetic therapies could help treat these tumors. 

The genome's gene count: still climbing

A key challenge remaining in deciphering the human genome is precisely identifying all of the regions that contain protein-coding sequences. Using the machine-learning tool PhyloCSF, which compares the genomes of multiple species to predict functional, conserved protein-coding sequences, a team led by Irwin Jungreis, associate member Manolis Kellis, and colleagues reports new insights into protein-coding DNA from human, mouse, chicken, fly, worm, and mosquito genomes. Among their discoveries, the researchers report 70 previously undetected protein-coding genes and reveal that 118 GWAS variants previously thought to be noncoding are in fact protein-altering. Read the full story in Genome Research.

Now screening: image-based phenotyping

Scientists use genetic screens to perturb genes in mammalian cells to learn what those genes do. Pooled screens take this same approach and can routinely scale to analyze many more genetic perturbations, but have not been compatible with imaging, which allows for simultaneous high-resolution measurement of many cell features. David Feldman, Avtar Singh, core institute member Paul Blainey, and colleagues have developed a new technology that combines large-scale pooled genetic screens with image-based phenotyping to examine how genes affect complex cellular processes, and are continuing to develop and apply the method through collaborations with the Cancer Program, Imaging Platform, Klarman Cell Observatory, and Stanley Center. Read more in Cell and a Broad news story.

To learn more about research conducted at the Broad, visit, and keep an eye on