Sneha D. Goenka joins Princeton faculty with expertise in computer systems architecture for advanced genomic applications

Written by
Alaina O'Regan
Jan. 22, 2025

Sneha D. Goenka, an expert in computer architecture and systems for computational genomics, has joined Princeton University as an assistant professor of electrical and computer engineering.

Goenka specializes in improving computing speed and scalability to tackle pivotal challenges in genomics, enabling more efficient healthcare diagnostics and evolutionary studies. As advancements in DNA sequencing have made genomics data more accessible, this rapid expansion has brought new computational challenges. “The bottleneck for genomics research has shifted from generating data to computing and analyzing that data,” Goenka said.

For instance, the first human genome, sequenced in 2003, took 13 years and $3 billion to complete. Today, the same process can be done in under a day for less than one thousand dollars. But computing capabilities haven’t kept pace to store and analyze all of this data. “While this advancement was happening in the genomics space, the performance improvements in the computing space kept sort of plateauing,” Goenka said.

Faster genomic analysis can be lifesaving, especially in cases like diagnosing genetic diseases in newborn infants in the ICU, according to Goenka. She designed systems that enabled the world-record setting genetic disease diagnosis, which could routinely provide the diagnosis in under eight hours. This is orders of magnitude faster than previous methods, which could take weeks, ultimately saving cost and lifting the burden of stress on the patient and their family. Goenka was awarded the Forbes 30 Under 30 award in the Science category for this work last year.  

At Princeton, Goenka plans to explore whether existing computing platforms can be repurposed to process large-scale genomic data. An example of this kind of adaptation is graphics processing chips, which were originally created to render 3D graphics, and today are relied upon for machine learning applications. “In a similar vein, I plan to look into how different aspects of computing could be used to make genomics research faster,” she said.

One of Goenka’s ongoing projects aims to reduce the memory footprint for genomic data by developing more efficient data compression techniques. Storing the raw signals for a single human genome uses on the order of two terabytes of data, according to Goenka. “We’re reaching a point where re-sequencing an entire genome is better cost-wise than storing the data,” she said.

Better compression techniques could lead to significant cost savings for research institutions and healthcare providers, making large-scale genomic projects more accessible.

This is also instrumental for cancer research, where storing and analyzing sequenced genomes from cancer cells requires vast amounts of time and storage. 

In addition to healthcare applications, faster genomic processing could enable scientists to glean insights into biodiversity and evolution on a much shorter timeline. Scientists sequence DNA to find out the genetic makeup of different species and better understand evolution. “A big obstacle is that we’re sequencing these species at a rate that is far outpacing our ability to analyze them,” Goenka said. She has already developed methods that increase the processing power of genomic analyses by 14 times, reducing the time it takes for some analyses from years to months.

Goenka is also working on co-designing algorithms, software and hardware systems to enable more efficient pan-genomic analysis, an emerging approach that aims to increase representation of genetic diversity in genomics research. Overcoming the computational challenges of this approach could lead to improved studies on disease susceptibility and personalized medicine.

Goenka holds a Ph.D. from Stanford University, as well as a bachelor’s and master’s degrees from the Indian Institute of Technology, Bombay. She completed her postdoctoral research at Stanford School of Medicine.