By Siddharth Vodnala, Voiland College of Engineering and Architecture
PULLMAN, Wash. – Washington State University researchers have received a $500,000 National Science Foundation grant to build technologies that can help process very large amounts of biological data.
Led by Ananth Kalyanaraman and Partha Pande, professors in the School of Electrical Engineering and Computer Science, the research could someday help to answer fundamental questions in biology, particularly in the important area of biological networks.
“Vast amounts of biological data are being generated using a variety of high throughput technologies,” said Kalyanaraman. “We are trying to create algorithmic techniques to first network models for such data, and subsequently process those networks faster and more efficiently.”
As part of the grant, the researchers are designing the software and hardware together, which they believe will help to optimize the technology.
On the software side, Kalyanaraman will design algorithms to process large data sets using parallel computing, which is a type of computing in which computer instructions are carried out simultaneously. His algorithms will be able to perform several mathematical operations on very large amounts of biological data.
For example, gut bacteria in humans play a very important role in human health. These bacteria can be thought of as an interconnected network. They often form tightly knit communities at different layers of the network. Finding such communities is crucial to our understanding of how such bacteria affect humans, but, because of the abundance of data, current software technology is unable to identify them.
Kalyanaraman will research new ways to speed up the processing by exploiting the connected features in the data, which collectively render themselves in the form of a network that can be subsequently processed. However, analyzing large network structures on a parallel platform poses significant design challenges which the team is planning to overcome.
“Traditionally, people developed algorithms without knowledge from chip architectures,” said Kalyanaraman. “Instead we are taking a co-design approach where algorithms and architectures can be co-designed.”
On the hardware side, Pande will be designing a computer chip that improves the processing speed for large amounts of data while also being energy efficient. Pande’s chip designs create heterogeneous platforms, comprising of both graphics processing units (GPUs) and central processing units (CPUs) to speed up data processing. This innovative design can reduce latency in operation, increase processing speed, and reduce power consumption.
At the end of the three-year grant period, the researchers plan to have a software library released to the community, that will allow scalable construction and analysis of biological networks on supercomputers.
The researchers also plan to complete the chip design aimed at speeding up data processing and reducing energy consumption by the end of the grant period.
- Brett Stav, public relations/marketing director, Voiland College of Engineering and Architecture, 509-335-8189, firstname.lastname@example.org