Partnering with NVIDIA Clara Parabricks to Accelerate NGS Data Processing

Looking for ways to accelerate data processing and genomic analysis? By using graphics processing units (GPUs), it’s possible to achieve throughput comparable to nearly 40-50 CPU servers with just one GPU server reducing overall operating costs.

We’re excited to be working with NVIDIA Clara Parabricks to make the DNA and RNA pipelines available on our DNAnexus Titan or Apollo Platforms. For a limited time, we are offering license-free access to test, validate, and use the following Clara Parabricks pipelines (more on the way):

GATK4 and DeepVariant users will appreciate NVIDIA Clara Parabricks Pipelines. They provide equivalent results, but at much greater throughput and speed — 50-90 times faster. In fact, Parabricks Pipelines v3.0 achieved a record for the germline analysis of a whole human genome in under 20 minutes when running on the just-announced NVIDIA A100 GPUs. On DNAnexus today, GATK analyses which take nearly 30 hours of computation on a 32-vCPU machine complete, will complete in 30 minutes with identical results on an eight v100 GPU instance using Parabricks.  The Parabricks Pipelines are compatible with GATK4.1 and DeepVariant 0.10.0, the latest major releases for both analysis packages.

The tools are also fully configurable, so users can choose to run the full GATK4 best practices, for instance, or specific steps from the pipeline.

DNAnexus makes it easy to run NVIDIA Parabricks pipelines by providing an out-of-the-box solution. Users simply supply input files and receive results delivered in their DNAnexus project. Launching NVIDIA Parabricks apps is a seamless process, removing the hassle of provisioning GPU instances in the cloud manually or installing CUDA or NVIDIA Parabricks apps themselves.

For secondary analyses of sequencer-generated FASTQ files to variant call files (VCFs), we recommend you register today to take advantage of this limited-time, license-free opportunity to try out Parabricks available through June 30th