APPLICATIONS NEWS

Genome boost in Washington

7 December 2012



Collaboration between researchers at the University of Washington and Pico Computing has yielded up to 90-fold acceleration of Infernal, a software package commonly used to identify non-coding ribonucleic acid (ncRNA). Infernal can take upwards of a few weeks to complete on commodity CPUs. However, with Pico’s FPGA accelerated solution, this identification process has been reduced to less than a day, resulting in an up to 90-fold improvement.

In the traditional model of molecular biology, DNA is transcribed to form RNA, which in turn is translated into proteins. These proteins perform many of the functions essential to biological life. In recent decades, biologists have come to understand the importance of ncRNAs, which directly perform roles normally associated with proteins including regulating genes and catalysing reactions. Further study of ncRNA may result in breakthroughs in the areas such as cancer, Alzheimer’s, and Parkinson’s disease. 

In the past 10 years, the number of modelled ncRNA families identified has increased by two orders of magnitude. Many ncRNA's bases have purely structural roles requiring two potentially distant bases to be complementary. This makes the search for ncRNA more difficult than traditional DNA sequence matching. Growth of these ncRNA families as well as computational complexity of searching genomes for known ncRNA has resulted in runtimes on the order of weeks.

Researchers at the University of Washington, using Pico’s M-503 FPGA module, have accelerated algorithms within Infernal. This implementation yielded individual algorithm speedups of up to 200-fold, for an overall software acceleration of up to 90-fold.

'The last decade of genome research has yielded a flood of novel non-protein-coding RNAs (ncRNAs) with diverse biological functions, and tantalizing hints of thousands more. Pico's hardware has enabled us to reach new levels of acceleration for key computationally intensive algorithms needed to fully explore this important new landscape,' said Dr Walter L. Ruzzo, professor at the university.

Related internet links

Pico Computing
University of Washington