Duke University is the recipient of two large grants totaling nearly $12 million from the National Human Genome Research Institute (NHGRI) as part of its new Impact of Genomic Variation on Function (IGVF) Consortium. The IGVF Consortium will bring investigators together in a highly collaborative effort to examine how genomes function, how genome function shapes phenotypes, and how these processes are influenced by genomic variation. Grant recipients are distributed across five new types of centers: Functional Characterization Centers, Regulatory Network Projects, Mapping Centers, Predictive Modeling Centers and a Data and Administrative Coordinating Center. NHGRI has committed a total of $185 million to the new IGVF Consortium across 25 awards, two of which are allocated to Duke, including both a Functional Characterization Center and a Predictive Modeling Center.
Duke IGVF Characterization Center
An $8.6 million award will fund the Duke Characterization Center, led by Charlie Gersbach, Greg Crawford and Tim Reddy. They are one of eight such centers, which serve as the main “factories” to generate data that will be used to inform how genetic variation impacts gene regulation and, ultimately, phenotypes. “We and others have identified millions of gene regulatory elements across the genome,” Crawford said. “However, we know very little about what almost all of them do.”
The goal of the Duke IGVF Functional Characterization Center is to systematically perturb large numbers of regulatory elements and study their function across different biological contexts. “It’s a critical next step to more fully understand how the genome works,” Crawford said.
Gene regulatory elements control almost all complex human traits. “We know that these regulatory elements are altered in complex diseases, define the identity of different cell types in the body, and modulate tissue regeneration and homeostasis,” Gersbach said. “We hope that assigning functional roles to many of these elements will also inform these processes.”
To do this, they will integrate several high throughput-based CRISPR technologies they have already developed along with single cell RNA sequencing, cell programming tools, transgenic mice, and a population-based high throughput reporter assay called POP-STARR to characterize the impact of noncoding genetic variation across the entire genome. “This center is a direct extension of what we set out to do with founding the Center for Advanced Genomic Technologies,” Gersbach said. “It also has direct relevance to the Center for Statistical Genetics and the Duke Center for Combinatorial Gene Regulation,” the latter of which is an NIH Center of Excellence in Genome Science (CEGS) funded center at Duke.
The team will make all of their data immediately available and they will share protocols, reagents and analysis tools to the scientific community.
Duke IGVF Predictive Modeling Center
A $3.2 million award will fund the Duke Predictive Modeling Center, led by Andrew Allen, David Page, and Sayan Mukherjee with ample support from Raluca Gordan, Bill Majoros and administrative support from Shannon Clarke. They are one of seven such centers. They will take the raw data generated in the Functional Characterization Centers and transform it into consumable information that predicts the effect of variation on function.
This work will be focused into three parts. First, they will simulate the data being developed in the Functional Characterization Centers and develop a simulation framework that can replicate the data. This will help us understand how to fine-tune the experiments in the Characterization Centers,” Allen said, “which will help us maximize the information that comes out of those experiments.”
The team will also develop new graphical, model-based machine learning approaches that predict the functional effect of noncoding variation on function in diverse cell types. “We hope to expand those approaches so we can start integrating large collections of data in a meaningful way,” Allen said.
Finally, the team will use population genetics to look for genetic elements in the population that affect function. “We think that new variants that do something important are not tolerated in an organism and can lead to disease,” Allen said. “We can use population genetics to see if a variant is likely to be disease relevant.”
Most of the time variants may not cause any impacts in the genome, but if researchers can figure out what variants do have a functional impact, they can begin to better understand how it affects disease. This will allow doctors to better diagnose and treat diseases.
“The IGVF framework is trying to understand functions in the non-coding sequences of the entire genome, so we can interpret and prioritize all variation in patients.” Allen said, “We need to know what types of variants to look for, not just for patient diagnosis but also for disease changing discoveries.”
National Impact of Duke Genomics
The awarding of two new Duke Centers in this nascent NIH IGVF Consortium is one of several recent examples of Duke’s growing international impact in genome sciences and engineering. Duke’s significant involvement in the IGVF follows its recognition as a NIH Center of Excellence in Genome Sciences just last year, one of only ten such Centers of Excellence in the country. Similarly, Duke was awarded three grants as part of the $190M NIH Somatic Cell Genome Editing Consortium, one of only four institutions in the country to receive three awards under that initiative.