Parabricks speeds up genome analysis with NVIDIA GPUs

Source: NVIDIA blog. Parabricks speeds up genome analysis
with NVIDIA GPUs.

Computing power is often measured by how quickly a genome can be sequenced. Today, the bottleneck for genomic insights is the computational analysis that follows sequencing. This detects key markers and outliers, called variants, in the genetic data.

US-based Parabricks, a startup and an NVIDIA Inception member, is shrinking the time this analysis takes from a couple of days to under an hour with the help of NVIDIA's graphics processing units (GPUs). “It’s the first application for secondary analysis of genomic data on a GPU, and it fully matches the state-of-the-art analytical pipeline,” said Dave Gregorka, President of Parabricks.

Ankit Sethia, Parabricks cofounder and Technical Lead, says the amount of genetic data generated is doubling almost every year. At around 300 GB per human genome sample, the computational demand is high. “When there are tens of thousands of patients, tens of thousands of samples you need to analyse, it can take years using CPU-based processing,” he said.

“By analysing much, much faster, you can get to the right problem and the right solution much more quickly,” said Sethia. He explained that processing genomic data is similar to processing an image for a GPU. Just as each pixel can be processed and calculated independently, so can genetic information. 

Parabricks' GenomeBricks software uses GPUs to rapidly analyse the genome. Running on a single NVIDIA DGX-1 server, Parabricks’ software can process more than 12,000 whole genomes per year — a feat that would require 40 CPU servers. For its genomic analysis, the startup uses NVIDIA CUDA, including the cuDNN deep learning libraries, as well as TensorRT inferencing software.

The results from the Parabricks software enable researchers to analyse trends in genomic data from entire populations more quickly, benefitting fields such as personalised medicine, drug discovery and disease treatment. Mutations and variants in the data can be identified fast, helping medical professionals to understand the patient at a genetic level and decide on a custom path for treatment.

Parabricks rolled out the initial version of its GenomeBricks software suite to select customers in March and it is already in use in Singapore, Japan and Thailand. The company is also working on large population study projects, including national initiatives for precision medicine.