NVIDIA Clara Parabricks

NVIDIA Clara™ Parabricks is a computational framework supporting genomics applications from DNA to RNA. It employs NVIDIA’s CUDA, HPC, AI, and data analytics stacks to build GPU accelerated libraries, pipelines, and reference application workflows for primary, secondary, and tertiary analysis. Clara Parabricks is a complete portfolio of off-the-shelf solutions coupled with a toolkit to support new application development to address the needs of genomic labs.




The Clara Parabricks Toolkit is a technology stack of CUDA accelerated libraries and deep learning modules, C++ and Python APIs, reference applications and integrations with 3rd party applications and workflows for high performance computing, deep learning, and data analytics tools in genomics.

Use the Clara Parabricks Toolkit to develop AI-assisted workflows, to optimize mapping, aligning, and polishing for de novo genome assembly, and enhance the resolution of single cell epigenomics.

Software Overview

NVIDIA CLARA PARABRICKS TOOLKIT

Deep Learning & Long Read Sequencing

Libraries - Python APIs

  • CUDA Mapper - CUDA accelerated all-to-all mapping of sequencing reads, used for genome assembly workflows.
  • CUDA Aligner - CUDA based library with accelerated algorithms for sequencing read alignment, used for genome assembly applications such as Racon and for variant calling.
  • CUDA POA - CUDA library for accelerated partial order alignment, used for genome assembly and basecalling.

Reference Applications

  • Atac-Seq Deep Learning Denoising - a deep learning application to improve coverage track denoising and peak calling from low-coverage or low-quality ATAC-Seq data. Outputs are in the standard file format.
  • RNA-Seq Analytics - an interactive notebook for single cell RNA-Seq data.
  • DL Variant Caller - a deep learning based variant caller that outputs in the standard file format.
  • Long Read Mapping - based on the minimap2 software, this application maps long sequencing read data and outputs in the standard file format.

3rd Party Applications

  • Racon - consensus application for de novo genome assembly that utilizes cudaAligner for accelerated alignment and cudaPOA for accelerated polishing.
  • Raven - application for de novo genome assembly of long uncorrected reads that utilizes cudaAligner for accelerated alignment and cudaPOA for accelerated polishing.
  • Bonito - basecaller for Oxford Nanopore reads that utilizes cudaPOA for accelerated consensus.
  • Medaka - a deep learning consensus application for de novo genome assembly.

Access the full Clara Parabricks Toolkit

Download on GitHub




Clara Parabricks Toolkit Resources

Get Questions Answered On Our Developer Forum

Connect with members of the community through our Developer Forum.




Post Your Question

De Novo Genome Assembly from Long Uncorrected Reads

Racon, a consensus module for genome assembly, enables consensus genomes with similar or better quality than state-of-the-art methods with greater speed.


Read Paper

Using the Clara Parabricks Toolkit Against COVID-19

Use this get started guide for long read basecalling, short read alignment, and long read polishing on github.



View in GitHub

Based on the Broad Institute’s Genome Analysis Toolkit (GATK), Clara Parabricks Pipelines enable GPU-accelerated GATK along with other third party tools, like Google’s DeepVariant caller. Currently, GATK v4.1 is supported and Best Practices are enabled. Starting with DNA sequencing reads, Clara Parabricks Pipelines map, align, filter and call variants for either germline or somatic variant detection. For RNA based projects, both STAR and STAR-Fusion align sequencing reads allowing for reads to be split to account for exon/intron boundaries, followed by variant calling.

Clara Parabricks Pipelines were built to optimize acceleration, accuracy and scalability. Users can achieve a 35-50X acceleration and 99.99% accuracy for variant calling when comparing against CPU-only BWA-GATK4 pipelines. It can run the full GATK4 Best Practices, and is also fully configurable. As a result, you can choose which steps, parameter settings, and versions of the pipeline to run.

Learn More



Join the NVIDIA Developer Program

Join Now