Ritambhara Singh Wins An NHGRI Genomic Innovator Award For ML Approaches To Reveal Gene Regulation Mechanisms In Diseases

Click the links that follow for more news about Ritambhara Singh and other recent accomplishments by our faculty.

Professor Ritambhara Singh of Brown CS and Brown University’s Center for Computational Molecular Biology has just received the National Human Genome Research Institute (NHGRI)’s Genomic Innovator Award, a highly selective honor for early career scientists. Part of the National Institutes of Health (NIH), NHGRI is bestowing the honor on eleven researchers in the field of genomics this year. 

The Genomic Innovator Award was developed to support innovative work by genomics investigators who are early in their careers and part of consortia or other team-science efforts. Now in its third year, it supports researchers whose works span various areas of genomics, including CRISPR technologies, brain-related disorders, single-cell genomics, and precision medicine. Unlike more traditional grants, which fund defined research projects, the Genomic Innovator Award provides the researcher with flexibility to pursue innovative research directions in a nimble fashion within a broad scientific area. 

“I’m excited about continuing and extending our data integration efforts in the lab with this award,” Ritambhara says. “In the era of big data, combining different types of genomic datasets in a way that provides us with important biological insights is a challenging task that we are very keen to tackle in the lab.”

Singh’s interdisciplinary research program’s primary goal is to develop machine learning approaches to reveal gene regulation mechanisms in diseases from the data through integration and interpretation. “Our current understanding of gene regulation,” she says, “is akin to solving a jigsaw puzzle. While many factors governing gene expression have been identified, how these ‘parts’ are pieced together to function as a whole remains unclear. My research has developed and applied state-of-the-art machine learning methods on genomics datasets to attempt to put together these pieces from the data.”

The primary objectives of her proposal include:

  1. applying and refining deep learning architectures that capture the underlying structure  in the data to integrate multi-omics datasets and connect them to gene expression via the prediction task

  2. developing interpretation methods for these models to specifically pick out signals that correlate with misregulated genes in disease cell lines, and 

  3. extending these frameworks to single-cell datasets to understand gene regulation at the single-cell resolution.

The outcome of the proposed projects will address the fundamental questions in gene regulation studies, specifically revealing the importance of gene regulatory mechanisms in characterizing diseases from the collected data.

“Our efforts,” Ritambhara says, “will lead to deep learning frameworks that integrate information and produce interpretable outcomes capturing how gene regulation is affected on a broad cell population and single-cell scale. It’s part of what I see as the ultimate goal for the future of genomics, which is to have a streamlined flow of information from the genomic data collection effort to our understanding of diseases to the clinical applications. I hope we continue to build collaborations and leverage methodological innovations to accomplish this goal.”

For more information, click the link that follows to contact Brown CS Communication and Outreach Specialist Jesse C. Polhemus.