Wikipedia

NVIDIA Parabricks

NVIDIA Parabricks is a suite of free software for genome analysis developed by Nvidia, designed to deliver high throughput by resorting to Graphics Processing Unit (GPU) acceleration.[1]

NVIDIA Parabricks
Developer(s)Nvidia
Stable release
v4.3.1-1 / July 1, 2024
PlatformNvidia GPUs
Available inEnglish
TypeHealth software
Websitehttps://www.nvidia.com/en-us/clara/genomics/

Parabricks offers workflows for DNA and RNA analyses and the detection of germline and somatic mutations, using open-source tools.[1] It is designed to improve the computing time of genomic data analysis while maintaining the flexibility required for various bioinformatics experiments.[1] Along with the speed of GPU-based processing, Parabricks ensures high accuracy, compliance with standard genomic formats and the ability to scale in order to handle very large datasets.[1]

Users can download and run Parabricks pipelines locally or directly deploy them on cloud providers, such as Amazon Web Services, Google Cloud, Oracle Cloud Infrastructure, and Microsoft Azure.[1]

Accelerated genome analysis fundamentals

edit
 
Standard pipeline to extract variants from an individual's genome
 
Sequencing machines able to identify the sequence of bases constituting the DNA

The massive reduction in sequencing costs[2] resulted in a significant increase in the size and the availability of genomics data[3] with the potential of revolutionizing many fields, from medicine to drug design.[4]

Starting from a biological sample (e.g., saliva or blood), it is possible to extract the individual's DNA and sequence it with sequencing machinery to translate the biological information into a textual sequence of bases.[5] Then, once the entire genome is obtained through the genome assembly process, the DNA can be analyzed to extract information that is key in several domains, including personalized medicine and medical diagnostics.[6]

Typically, genomics data analysis is performed with tools based on Central Processing Units (CPUs) for processing.[7] Recently, several researchers in this field have underlined the challenges in terms of computing power delivered by these tools and focused their efforts on finding ways to boost the performance of the applications.[7] The issue has been addressed in two ways: developing more efficient algorithms or accelerating the compute-intensive part using hardware accelerators. Examples of accelerators used in the domain are GPUs, FPGAs, and ASICs[8]

In this context, GPUs have revolutionized genomics by exploiting their parallel processing power to accelerate computationally intensive tasks.[9][10] GPUs deliver promising results in these scenarios thanks to their architecture, composed of thousands of small cores capable of performing computations in parallel.[11] This parallelism allows GPUs to process multiple tasks simultaneously, significantly speeding up computations that can be broken down into independent units.[11] For instance, aligning millions of sequencing reads against a reference genome or performing statistical analyses on large genomic datasets can be completed much faster on GPUs than when using CPUs.[10] This facilitates the rapid analysis of genomic data from diverse sources, ranging from individual genomes to large-scale population studies,[12] accelerating the understanding of genetic diseases, genetic diversity, and more complex biological systems.[10]

edit

Parabricks offers end users various collections of tools organized sequentially to analyze the raw data according to the user's requirements, called pipelines.[1] Nevertheless, users can decide to run the tools provided by Parabricks as a standalone, still exploiting GPU acceleration to overcome possible computational bottlenecks. Only some of the provided tools in the suite are GPU-based.[13]

 
Overview of the main steps of NVIDIA Parabricks pipelines

Overall, all the pipelines share a standard structure. Most of the pipelines are built to analyze FASTQ data resulting from various sequencing technologies (e.g., short- or long-read). Input genomic sequences are firstly aligned and then undergo a quality control process. These two processes provide a BAM or a CRAM file as an intermediate result. Based on this data, the variant calling task that follows employs high-accuracy tools that are already widely used. As output, these pipelines provide the identified mutations in a VCF (or a gVCF).[13]

Germline pipeline

edit

The germline pipeline offered by Parabricks follows the best practices[14] proposed by the Broad Institute in their Genome Analysis ToolKit (GATK).[15] The germline pipeline operates on the FASTQ files provided as input by the user to call the variants that, belonging to the germ line, can be inherited.[13]

This pipeline analyzes data computing the read alignment with BWA-MEM[16][17] and calling variants using GATK HaplotypeCaller,[18] one of the most relevant tools in the domain for germline variant calling.[13]

DeepVariant germline pipeline

edit

Besides the pipeline that resorts to HaplotypeCaller to call variants, Parabricks also offers an alternative pipeline that still calls germline variants but is based on DeepVariant.[19][20] DeepVariant is a variant caller, developed and maintained by Google, capable of identifying mutations using a deep learning-based approach. The core of DeepVariant[19] is a convolutional neural network (CNN) that identifies variants by transforming this task into an image classification operation. In Parabricks, the inference process is accelerated in hardware. For this pipeline, only T4, V100, and A100 GPUs are supported.[13]

Analyses performed according to this pipeline are compliant with the use of BWA-MEM[16] for the alignment by Google's CNN for variant calling.[13]

Human_par pipeline

edit

Still compliant with GATK best practices,[14] the human_par pipeline allows users to identify mutations in the entire human genome, including sex chromosomes X and Y, and, thus, it is compliant with their ploidy. For male samples, firstly, the pipeline runs HaplotypeCaller[18] on all the regions that do not belong to the X and Y chromosomes and on the pseudoautosomal region with ploidy equal to 1. Then, HaplotypeCaller analyses the X and Y regions without the pseudoautosomal region with ploidy 2. Regarding female samples, instead, the pipeline runs HaplotypeCaller on the entire genome, with ploidy 2.[13]

The sex of the sample can be determined in two main ways:

  1. Manually set with the --sample-sex option;
  2. Specify the X vs. Y ratio with range options --range-male and --range-female and let the tool automatically infer the sex of the samples based on the X and Y reads count.

The pipeline requires the user to specify at least one of these three options.[13]

As for the germline case, since this pipeline targets the germline variants, the pipeline resorts to BWA-MEM[16] for the alignment, followed by HaplotypeCaller[18] for variant calling.[13]

Somatic pipeline

edit

Parabricks' somatic pipeline is designed to call somatic variants, i.e., those mutations affecting non-reproductive (somatic) cells. This pipeline can analyze both tumor and non-tumor genomes, offering either tumor-only or tumor/normal analyses for comprehensive examinations.[13]

As in the germline pipeline, the alignment task is carried out using BWA-MEM[16] followed by GATK Mutect[21] to identify the possible mutations. Mutect is used instead of HaplotypeCaller due to its focus on somatic mutations, as opposed to germline mutations targeted by HaplotypeCaller.[21]

RNA pipeline

edit

This pipeline is optimized for short variant discovery (i.e., single-nucleotide polymorphisms (SNPs) and indels) in RNAseq data. It follows the Broad Institute's best practices for these types of analyses.[13]

It relies on the STAR aligner,[22] a read aligner specialized for RNA sequences for aligning the reads, and HaplotypeCaller[18] for calling variants.[13]

Parabricks tools

edit

Parabricks provides a collection of tools to perform genomics analyses, classified into six main categories related to their task.[13] These tools combined constitutes Parabricks' pipelines, and can be also used as-is.

For FASTQ and BAM files processing, the proposed tools are[13]:

  • applybsqr
  • bam2fq
  • bamsort
  • bqsr
  • fq2bam
  • fq2bamfast
  • fq2bam_meth
  • markdup
  • minimap2 (beta)

For calling variants, the proposed tools are[13]:

  • deepsomatic
  • deepvariant
  • deepvariant_germline
  • germline (GATK Germline Pipeline)
  • haplotypecaller
  • mutectcaller
  • pacbio_germline (beta)
  • postpon
  • prepon
  • somatic (Somatic Variant Caller)

For RNA processing, the proposed tools are[13]:

  • rna_fq2bam
  • starfusion

For results quality control, the proposed tools are[13]:

  • bammetrics
  • collectmultiplemetrics

For processing variants, the proposed tools are[13]:

  • dbsnp

For processing gVCF files, the proposed tools are[13]:

  • genotypegvcf
  • indexgvcf

Not all the listed tools are accelerated on GPU.[13]

Hardware support

edit

Users can download and run Parabricks pipelines on their local servers, allowing for private, on-site data processing and analysis. They also can deploy Parabricks pipelines on cloud platforms, with improved scalability for larger datasets. Supported cloud providers include AWS, GCP, OCI, and Azure.[1]

In the latest release (v4.3.1-1), Parabricks includes support for the NVIDIA Grace Hopper super chip.[23] The NVIDIA GH200 Grace Hopper Superchip is a heterogeneous platform designed for high-performance computing and artificial intelligence, combining an NVIDIA Grace and a Hopper on a single chip.[24] This platform enhances application performance using both GPUs and CPUs, offering a programming model aimed at improving performance, portability, and productivity.[23]

Applications

edit

Due to the computational power required by genomics workloads, Parabricks has found application in several research studies with different applicative domains, especially in cancer research.[25][26][27]

Scientists from Washington University used the Parabricks DeepVariant pipeline for identifying variants (e.g., SNPs and small indels) in long-read Hi-Fi whole-genome sequencing (WGS) data generated with PacBio's Revio SMRT Cell technology.[28]

In addition to the pipelines, individual components of Parabricks have been used as standalone tools in academic settings. For example, the accelerated DeepVariant has been employed in a novel process to reduce the processing time further for WGS Nanopore data.[29]

In 2022, Nvidia announced a collaboration with the Broad Institute to provide researchers with the benefits of accelerated computing. This partnership includes the entire suite of Nvidia's biomedical hardware-accelerated software suite called Clara, that includes Parabricks and MONAI.[30] Similarly, the Regeneron Genetics Center uses Parabricks to expedite the secondary analysis of the exomes they sequence in their high-throughput sequencing center, leverage the DeepVariant Germline pipeline inside their workflows.[31]

See also

edit

References

edit
  1. ^ a b c d e f g "Clara for Genomics". NVIDIA. Retrieved 8 July 2024.
  2. ^ "DNA Sequencing Costs: Data". www.genome.gov. Retrieved 2024-07-10.
  3. ^ Langmead B, Nellore A (April 2018). "Cloud computing for genomic data analysis and collaboration". Nature Reviews. Genetics. 19 (4): 208–219. doi:10.1038/nrg.2017.113. PMC 6452449. PMID 29379135.
  4. ^ Ombrello MJ, Sikora KA, Kastner DL (April 2014). "Genetics, genomics, and their relevance to pathology and therapy". Best Practice & Research. Clinical Rheumatology. Advances in Paediatric Rheumatology and Translation of Research to Targeted Therapies. 28 (2): 175–189. doi:10.1016/j.berh.2014.05.001. PMC 4149217. PMID 24974057.
  5. ^ Alser M, Lindegger J, Firtina C, Almadhoun N, Mao H, Singh G, et al. (2022). "From molecules to genomic variations: Accelerating genome analysis via intelligent algorithms and architectures". Computational and Structural Biotechnology Journal. 20: 4579–4599. doi:10.1016/j.csbj.2022.08.019. PMC 9436709. PMID 36090814.
  6. ^ Jain KK (2009). "Basics of Personalized Medicine". In Jain KK (ed.). Textbook of Personalized Medicine. New York, NY: Springer. pp. 1–27. doi:10.1007/978-1-4419-0769-1_1. ISBN 978-1-4419-0769-1.
  7. ^ a b Alser M, Bingol Z, Cali DS, Kim J, Ghose S, Alkan C, et al. (September 2020). "Accelerating Genome Analysis: A Primer on an Ongoing Journey". IEEE Micro. 40 (5): 65–75. arXiv:2008.00961. doi:10.1109/MM.2020.3013728. ISSN 0272-1732.
  8. ^ Alser M, Rotman J, Deshpande D, Taraszka K, Shi H, Baykal PI, et al. (August 2021). "Technology dictates algorithms: recent developments in read alignment". Genome Biology. 22 (1): 249. doi:10.1186/s13059-021-02443-7. PMC 8390189. PMID 34446078.
  9. ^ Taylor-Weiner A, Aguet F, Haradhvala NJ, Gosai S, Anand S, Kim J, et al. (November 2019). "Scaling computational genomics to millions of individuals with GPUs". Genome Biology. 20 (1): 228. doi:10.1186/s13059-019-1836-7. PMC 6823959. PMID 31675989.
  10. ^ a b c Nobile MS, Cazzaniga P, Tangherloni A, Besozzi D (September 2017). "Graphics processing units in bioinformatics, computational biology and systems biology". Briefings in Bioinformatics. 18 (5): 870–885. doi:10.1093/bib/bbw058. PMC 5862309. PMID 27402792.
  11. ^ a b Cheng J, Grossman M, McKercher T (2014-09-09). Professional CUDA C Programming. John Wiley & Sons. ISBN 978-1-118-73932-7.
  12. ^ Zhou C, Lang X, Wang Y, Zhu C (2015-08-06). "gPGA: GPU Accelerated Population Genetics Analyses". PLOS ONE. 10 (8): e0135028. Bibcode:2015PLoSO..1035028Z. doi:10.1371/journal.pone.0135028. PMC 4527771. PMID 26248314.
  13. ^ a b c d e f g h i j k l m n o p q r s t "Welcome to NVIDIA Parabricks v4.3.1". NVIDIA Docs. Retrieved 2024-07-10.
  14. ^ a b "Best Practices for Variant Calling with the GATK". @broadinstitute. 2015-03-19. Retrieved 2024-07-09.
  15. ^ "Genome Analysis Toolkit (GATK)". @broadinstitute. 2010-06-08. Retrieved 2024-07-09.
  16. ^ a b c d Li H (2013-05-26), Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM, arXiv:1303.3997
  17. ^ "Burrows-Wheeler Aligner". bio-bwa.sourceforge.net. Retrieved 2024-07-09.
  18. ^ a b c d Poplin R, Ruano-Rubio V, DePristo MA, Fennell TJ, Carneiro MO, Van der Auwera GA, et al. (2018-07-24), Scaling accurate genetic variant discovery to tens of thousands of samples, doi:10.1101/201178, retrieved 2024-07-09
  19. ^ a b Poplin R, Chang PC, Alexander D, Schwartz S, Colthurst T, Ku A, et al. (November 2018). "A universal SNP and small-indel variant caller using deep neural networks". Nature Biotechnology. 36 (10): 983–987. doi:10.1038/nbt.4235. PMID 30247488.
  20. ^ google/deepvariant, Google, 2024-07-04, retrieved 2024-07-09
  21. ^ a b Cibulskis K, Lawrence MS, Carter SL, Sivachenko A, Jaffe D, Sougnez C, et al. (March 2013). "Sensitive detection of somatic point mutations in impure and heterogeneous cancer samples". Nature Biotechnology. 31 (3): 213–219. doi:10.1038/nbt.2514. PMC 3833702. PMID 23396013.
  22. ^ Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. (January 2013). "STAR: ultrafast universal RNA-seq aligner". Bioinformatics. 29 (1): 15–21. doi:10.1093/bioinformatics/bts635. PMC 3530905. PMID 23104886.
  23. ^ a b "Grace Hopper Superchip". NVIDIA Docs. Retrieved 2024-07-10.
  24. ^ Simakov NA, Jones MD, Furlani TR, Siegmann E, Harrison RJ (2024-01-11). "First Impressions of the NVIDIA Grace CPU Superchip and NVIDIA Grace Hopper Superchip for Scientific Workloads". Proceedings of the International Conference on High Performance Computing in Asia-Pacific Region Workshops. HPCAsia '24 Workshops. New York, NY, USA: Association for Computing Machinery. pp. 36–44. doi:10.1145/3636480.3637097. ISBN 979-8-4007-1652-2.
  25. ^ Crowgey EL, Vats P, Franke K, Burnett G, Sethia A, Harkins T, et al. (July 2021). "Abstract 165: Enhanced processing of genomic sequencing data for pediatric cancers: GPUs and machine learning techniques for variant detection". Cancer Research. 81 (13_Supplement): 165. doi:10.1158/1538-7445.AM2021-165. ISSN 0008-5472.
  26. ^ Ng JK, Vats P, Fritz-Waters E, Sarkar S, Sams EI, Padhi EM, et al. (December 2022). "de novo variant calling identifies cancer mutation signatures in the 1000 Genomes Project". Human Mutation. 43 (12): 1979–1993. doi:10.1002/humu.24455. PMC 9771978. PMID 36054329.
  27. ^ Lee TH, Jang BS, Chang JH, Kim E, Park JH, Chie EK (July 2023). "Genomic landscape of locally advanced rectal adenocarcinoma: Comparison between before and after neoadjuvant chemoradiation and effects of genetic biomarkers on clinical outcomes and tumor response". Cancer Medicine. 12 (14): 15664–15675. doi:10.1002/cam4.6169. PMC 10417181. PMID 37260182.
  28. ^ Manuel JG, Heins HB, Crocker S, Neidich JA, Sadzewicz L, Tallon L, et al. (June 2023). "High Coverage Highly Accurate Long-Read Sequencing of a Mouse Neuronal Cell Line Using the PacBio Revio Sequencer". bioRxiv. doi:10.1101/2023.06.06.543940. PMC 10274723. PMID 37333171.
  29. ^ Goenka SD, Gorzynski JE, Shafin K, Fisk DG, Pesout T, Jensen TD, et al. (July 2022). "Accelerated identification of disease-causing variants with ultra-rapid nanopore genome sequencing". Nature Biotechnology. 40 (7): 1035–1041. doi:10.1038/s41587-022-01221-5. PMC 9287171. PMID 35347328.
  30. ^ "The Broad Institute and NVIDIA Bring NVIDIA Clara to Terra Cloud Platform Serving 25,000 Researchers Advancing Biomedical Discovery". NVIDIA Newsroom. Retrieved 2024-07-09.
  31. ^ "UK Biobank Advances Genomics Research with NVIDIA Clara Parabricks". NVIDIA. Retrieved 2024-07-09.

Further reading

edit
edit
Retrieved from "https://en.wikipedia.org/w/index.php?title=NVIDIA_Parabricks&oldid=1234137251"