Cost-effective and accurate genomics analysis with Sentieon on AWS

By Amazon Web Services

January 31, 2023

This blog post was contributed by Don Freed, Senior Bioinformatics Scientist, and Brendan Gallagher, Head of Business Development at Sentieon; and Olivia Choudhury, PhD, Senior Partner Solutions Architect, Sujaya Srinivasan, Genomics Solutions Architect, and Aniket Deshpande, Senior Specialist, HPC HCLS at AWS.

The year 2022 was an exciting one for genomics customers, with announcements from Illumina’s targeting $2/Gbase for 30X whole genomes, and new instruments from Element BiosciencesUltima Genomics, and Pacific Biosciences. While these platforms offer exciting new choices for genomic data generation, they also present new challenges for data analysis due to unique platform-specific data characteristics and the large amount of data generated. These challenges highlight the need for efficient and scalable solutions for genomic data processing that are flexible enough to handle the unique properties of these emerging platforms.

Sentieon develops highly-optimized algorithms for bioinformatics applications, providing accurate, efficient, performant, and robust tools for the analysis of next generation sequencing data. Sentieon DNAseq and TNseq provide an implementation of the GATK best practices pipelines for germline and somatic variant calling with improved robustness and performance. Sentieon’s DNAscope pipelines are shown to be highly performant on Illumina data while offering substantially improved germline variant calling accuracy compared to GATK best practices.

In this blog post, we benchmark the performance of Sentieon’s DNAseq and DNAscope pipelines using publicly available datasets from Illumina, PacBio HiFi, Element Biosciences, and Ultima Genomics platforms using Amazon Elastic Compute Cloud (Amazon EC2) instances. Readers will gain an understanding of the runtime, cost, and accuracy performance of state-of-the-art germline variant calling pipelines across a wide range of Amazon EC2 instances. These benchmarks use publicly available datasets and the pipeline is available on GitHub.

Benchmarking Sentieon’s DNAseq and DNAscope pipelines

Pre-requisites

  • AWS account with permission to provision Amazon EC2, Amazon EBS, and Amazon VPC.
  • Deployment of a Sentieon license server inside of your VPC. Sentieon provides a Deployment Guide for AWS at, https://support.sentieon.com/appnotes/aws_deployment/.
  • An EC2 instance within the same VPC as the Sentieon license server.

Running Sentieon on AWS

An implementation of the Sentieon pipeline used in this blog post along with scripts for instance setup are provided on GitHub. The command `bash instance_setup.sh` will download the appropriate build of the Sentieon software and install other dependencies for this benchmark workflow. The pipelines benchmarked in this blog can be run with `bash run_benchmarks.sh`. The pipelines will download the input datasets, reference, and model files, down-sample the input read data to a consistent number of bases, run the Sentieon secondary analysis pipelines with the input dataset, and calculate variant calling accuracy against the Genome in a Bottle (GIAB) v4.2.1 truth set.

In practice, we ran the first part of the pipeline once to prepare the files used in the benchmark and then stored the prepared files on an AWS FSx for Lustre file system. The preprocessed files were used in later benchmarks on other instance types.

An Amazon gp3 EBS volume with 10,000 IOPS and a throughput of 700 MB/s was used for intermediate file storage. We found that this EBS disk worked well with the demanding IO needs of the Sentieon pipelines on the largest instances. For benchmarks using the hpc6a.48xlarge instance, the EBS disk was supplemented with a ramdisk for faster IO operations.

Benchmark Analysis

In these benchmarks, we use the GIAB HG002 sample sequenced on multiple sequencing platforms. As described in the pipeline config file, input FASTQ files were obtained from the following URLs:

  • Illumina NovaSeq
    • https://s3.amazonaws.com/genomics-benchmark-datasets/google-brain/fastq/novaseq/wgs_pcr_free/40x/HG002.novaseq.pcr-free.40x.R1.fastq.gz
    • https://s3.amazonaws.com/genomics-benchmark-datasets/google-brain/fastq/novaseq/wgs_pcr_free/40x/HG002.novaseq.pcr-free.40x.R2.fastq.gz
  • Illumina HiSeq
    • https://s3.amazonaws.com/genomics-benchmark-datasets/google-brain/fastq/novaseq/wgs_pcr_free/40x/HG002.novaseq.pcr-free.40x.R1.fastq.gz
    • https://s3.amazonaws.com/genomics-benchmark-datasets/google-brain/fastq/novaseq/wgs_pcr_free/40x/HG002.novaseq.pcr-free.40x.R2.fastq.gz
  • Element Biosciences Aviti
    • Downloaded from https://go.elementbiosciences.com/access-seq-datasets-060622
  • Ultima Genomics UG100
    • https://s3.amazonaws.com/ultima-selected-1k-genomes/crams/005401-UGAv3-1-CACATCCTGCATGTGAT.cram
  • PacBio HiFi
    • https://s3.amazonaws.com/giab/data/AshkenazimTrio/HG002_NA24385_son
      /PacBio_CCS_15kb_20kb_chemistry2/reads/m64011_190830
      _220126.fastq.gz
    • https://s3.amazonaws.com/giab/data/AshkenazimTrio/HG002_NA24385_son
      /PacBio_CCS_15kb_20kb_chemistry2/reads/m64011_190901
      _095311.fastq.gz
    • https://s3.amazonaws.com/giab/data/AshkenazimTrio/HG002_NA24385_son
      /PacBio_CCS_15kb_20kb_chemistry2/reads/m64012_190920
      _173625.fastq.gz
    • https://s3.amazonaws.com/giab/data/AshkenazimTrio/HG002_NA24385_son
      /PacBio_CCS_15kb_20kb_chemistry2/reads/m64012_190921
      _234837.fastq.gz
    • https://s3.amazonaws.com/giab/data/AshkenazimTrio/HG002_NA24385_son
      /PacBio_CCS_15kb_20kb_chemistry2/reads/m64015_190920
      _185703.fastq.gz
    • https://s3.amazonaws.com/giab/data/AshkenazimTrio/HG002_NA24385_son
      /PacBio_CCS_15kb_20kb_chemistry2/reads/m64015_190922
      _010918.fastq.gz

The input files vary in their coverage, so the datasets with FASTQ input were down-sampled to approximately 93 billion bases (~30x coverage) prior to processing with the Sentieon secondary analysis pipelines. The Ultima CRAM file was not down-sampled and is at 40x coverage as recommended by Ultima Genomics.

The data were processed using the hg38 reference genome. The reference genome at https://giab.s3.amazonaws.com/release/references/GRCh38/GCA_000001405.15
_GRCh38_no_alt_analysis_set.fasta.gz was used for files with input in the FASTQ format. The reference genome at https://broad-references.s3.amazonaws.com/hg38/v0/Homo_sapiens_assembly38.fasta was used with the Ultima data in CRAM format, as this dataset was already aligned to this reference genome.

Benchmarking Sentieon’s variant calling pipelines with Hpc6a instances

The Sentieon software is highly flexible, and can be used to process data with a wide variety of EC2 instance types. The recently announced Hpc6a instance family offers competitive price performance for compute intensive genomic secondary analysis workloads and is a good fit for Sentieon’s secondary analysis pipelines. Figure 1 presents the runtime and On-Demand compute cost of running Sentieon’s secondary analysis pipelines for germline variant calling across multiple sequencing technologies on hpc6a.48xlarge instances in the US East (Ohio) Region at the time of publication.

Figure 1: Runtime and On-Demand compute cost of the Sentieon DNAseq and DNAscope pipelines on a hpc6a.48xlarge instance. The Illumina, Element, and PacBio pipelines benchmark data processing from FASTQ to VCF. The Ultima pipeline benchmarks performance from CRAM to VCF.

On the hpc6a.48xlarge instance, analyzing a 30x Illumina NovaSeq sample from FASTQ to VCF took Sentieon’s DNAseq pipeline data 32 minutes with an On-Demand compute cost of $1.53. Sentieon’s DNAscope pipeline reduces SNP errors by 53% and INDEL errors by 78% on the 30x Illumina NovaSeq dataset compared to DNAseq while taking only 3-5 minutes longer — about 35 to 38 minutes for the various short-read platforms. On-Demand cost for DNAseq and DNAscope pipelines was typically $1.75, give or take a few cents.

The Ultima UG100 dataset was already aligned to the reference genome and in this benchmark, we only perform variant calling without alignment. Variant calling (CRAM to VCF) with the Ultima dataset finished in under 22 minutes for $1.05 in On-Demand compute costs on hpc6a.48xlarge.

Sentieon’s DNAscope LongRead pipeline for PacBio HiFi data is more computationally intensive as it includes multiple passes of variant calling along with a read-backed phasing. The DNAscope LongRead pipeline finished in 77 minutes with an On-Demand cost of $3.71.

The Element Biosciences AVITI system is a new desktop sequencer that became commercially available in the Spring of 2022 and is supported by a customized Sentieon DNAscope pipeline. Sentieon’s DNAscope pipeline for Element Biosciences finished in 35 minutes with an On-Demand compute cost of $1.69.

A full breakdown of the run times and costs are in Table 1.

Table 1: Runtime and On-Demand compute cost of the stages of the Sentieon DNAseq and DNAscope pipelines. Alignment includes alignment with Sentieon BWA-MEM for short-read data and alignment with Sentieon minimap2 for PacBio HiFi data. Preprocessing includes duplicate marking, base-quality score recalibration, and merging of multiple aligned files into a single file. Variant calling includes variant calling or variant candidate identification along with variant genotyping and filtering. Variant calling for PacBio HiFi data is implemented as a multi-stage pipeline. All analysis were done on hpc6a.48xlarge instances in the US East (Ohio) Region and are accurate at the time of publishing this post.

Sample Pipeline Alignment (min) Preprocessing (min) Variant Calling (min) Total Runtime (min) On-Demand hourly rate ($/hr) On-Demand compute cost ($)
Element Aviti DNAscope 23.2 1.1 10.8 35.2 2.88 1.69
Illumina HiSeq X DNAseq 23.7 2.6 4.2 30.5 2.88 1.46
Illumina HiSeq X DNAscope 23.7 1.0 10.5 35.2 2.88 1.69
Illumina NovaSeq DNAseq 25.1 2.3 4.5 31.9 2.88 1.53
Illumina NovaSeq DNAscope 25.1 0.9 11.6 37.6 2.88 1.81
PacBio HiFi DNAscope 24.8 1.6 50.9 77.2 2.88 3.71
Ultima UG100 DNAscope N/A N/A 21.2 21.2 2.88 1.02

Variant calling accuracy of the Sentieon DNAseq and DNAscope pipelines

The variant calling accuracy of the Sentieon pipelines as measured against the GIAB v4.2.1 truthset for the HG002 sample is shown in Figure 2. Sentieon’s DNAscope provides highly accurate SNP and indel variant calling across all platforms, with SNP F1-scores exceeding 99.5% and indel F1-scores exceeding 99.2%. Sentieon’s DNAseq pipeline provides identical results to the GATK best practices pipeline for germline variant calling, but has lower accuracy on the GIAB benchmark dataset.

Figure 2: Variant calling accuracy of the Sentieon DNAseq and DNAscope pipelines for SNVs, measured with the GIAB v4.2.1 benchmark VCF and BED files for HG002. Sentieon DNAseq matches the GATK HaplotypeCaller, but has lower accuracy on the Genome in a Bottle benchmark.
Figure 3: Variant calling accuracy of the Sentieon DNAseq and DNAscope pipelines for indels, measured with the GIAB v4.2.1 benchmark VCF and BED files for HG002. Sentieon DNAseq matches the GATK HaplotypeCaller, but has lower accuracy on the Genome in a Bottle benchmark.

The Ultima Genomics UG100 platform generates 20-24 WGS samples per day with a single instrument using a novel sequencing chemistry. During each sequencing cycle, this chemistry may incorporate zero, one, or more bases and the measured signal strength is proportional to the number of incorporated bases. Ultima Genomics, in their initial publication with the Broad Institute, has defined a high-confidence region which excludes 1.8% of the GIAB v4.2.1 high-confidence region with long homopolymers and other difficult regions

Read the full blog to learn more. Reminder: You can learn a lot from AWS HPC engineers by subscribing to the HPC Tech Short YouTube channel, and following the AWS HPC Blog channel.

Return to Solution Channel Homepage
Subscribe to HPCwire's Weekly Update!

Be the most informed person in the room! Stay ahead of the tech trends with industry updates delivered to you every week!

Nvidia’s New Blackwell GPU Can Train AI Models with Trillions of Parameters

March 18, 2024

Nvidia's latest and fastest GPU, code-named Blackwell, is here and will underpin the company's AI plans this year. The chip offers performance improvements from its predecessors, including the red-hot H100 and A100 GPUs. Read more…

Nvidia Showcases Quantum Cloud, Expanding Quantum Portfolio at GTC24

March 18, 2024

Nvidia’s barrage of quantum news at GTC24 this week includes new products, signature collaborations, and a new Nvidia Quantum Cloud for quantum developers. While Nvidia may not spring to mind when thinking of the quant Read more…

2024 Winter Classic: Meet the HPE Mentors

March 18, 2024

The latest installment of the 2024 Winter Classic Studio Update Show features our interview with the HPE mentor team who introduced our student teams to the joys (and potential sorrows) of the HPL (LINPACK) and accompany Read more…

Houston We Have a Solution: Addressing the HPC and Tech Talent Gap

March 15, 2024

Generations of Houstonian teachers, counselors, and parents have either worked in the aerospace industry or know people who do - the prospect of entering the field was normalized for boys in 1969 when the Apollo 11 missi Read more…

Apple Buys DarwinAI Deepening its AI Push According to Report

March 14, 2024

Apple has purchased Canadian AI startup DarwinAI according to a Bloomberg report today. Apparently the deal was done early this year but still hasn’t been publicly announced according to the report. Apple is preparing Read more…

Survey of Rapid Training Methods for Neural Networks

March 14, 2024

Artificial neural networks are computing systems with interconnected layers that process and learn from data. During training, neural networks utilize optimization algorithms to iteratively refine their parameters until Read more…

Nvidia’s New Blackwell GPU Can Train AI Models with Trillions of Parameters

March 18, 2024

Nvidia's latest and fastest GPU, code-named Blackwell, is here and will underpin the company's AI plans this year. The chip offers performance improvements from Read more…

Nvidia Showcases Quantum Cloud, Expanding Quantum Portfolio at GTC24

March 18, 2024

Nvidia’s barrage of quantum news at GTC24 this week includes new products, signature collaborations, and a new Nvidia Quantum Cloud for quantum developers. Wh Read more…

Houston We Have a Solution: Addressing the HPC and Tech Talent Gap

March 15, 2024

Generations of Houstonian teachers, counselors, and parents have either worked in the aerospace industry or know people who do - the prospect of entering the fi Read more…

Survey of Rapid Training Methods for Neural Networks

March 14, 2024

Artificial neural networks are computing systems with interconnected layers that process and learn from data. During training, neural networks utilize optimizat Read more…

PASQAL Issues Roadmap to 10,000 Qubits in 2026 and Fault Tolerance in 2028

March 13, 2024

Paris-based PASQAL, a developer of neutral atom-based quantum computers, yesterday issued a roadmap for delivering systems with 10,000 physical qubits in 2026 a Read more…

India Is an AI Powerhouse Waiting to Happen, but Challenges Await

March 12, 2024

The Indian government is pushing full speed ahead to make the country an attractive technology base, especially in the hot fields of AI and semiconductors, but Read more…

Charles Tahan Exits National Quantum Coordination Office

March 12, 2024

(March 1, 2024) My first official day at the White House Office of Science and Technology Policy (OSTP) was June 15, 2020, during the depths of the COVID-19 loc Read more…

AI Bias In the Spotlight On International Women’s Day

March 11, 2024

What impact does AI bias have on women and girls? What can people do to increase female participation in the AI field? These are some of the questions the tech Read more…

Alibaba Shuts Down its Quantum Computing Effort

November 30, 2023

In case you missed it, China’s e-commerce giant Alibaba has shut down its quantum computing research effort. It’s not entirely clear what drove the change. Read more…

Nvidia H100: Are 550,000 GPUs Enough for This Year?

August 17, 2023

The GPU Squeeze continues to place a premium on Nvidia H100 GPUs. In a recent Financial Times article, Nvidia reports that it expects to ship 550,000 of its lat Read more…

Analyst Panel Says Take the Quantum Computing Plunge Now…

November 27, 2023

Should you start exploring quantum computing? Yes, said a panel of analysts convened at Tabor Communications HPC and AI on Wall Street conference earlier this y Read more…

Shutterstock 1285747942

AMD’s Horsepower-packed MI300X GPU Beats Nvidia’s Upcoming H200

December 7, 2023

AMD and Nvidia are locked in an AI performance battle – much like the gaming GPU performance clash the companies have waged for decades. AMD has claimed it Read more…

DoD Takes a Long View of Quantum Computing

December 19, 2023

Given the large sums tied to expensive weapon systems – think $100-million-plus per F-35 fighter – it’s easy to forget the U.S. Department of Defense is a Read more…

Synopsys Eats Ansys: Does HPC Get Indigestion?

February 8, 2024

Recently, it was announced that Synopsys is buying HPC tool developer Ansys. Started in Pittsburgh, Pa., in 1970 as Swanson Analysis Systems, Inc. (SASI) by John Swanson (and eventually renamed), Ansys serves the CAE (Computer Aided Engineering)/multiphysics engineering simulation market. Read more…

Intel’s Server and PC Chip Development Will Blur After 2025

January 15, 2024

Intel's dealing with much more than chip rivals breathing down its neck; it is simultaneously integrating a bevy of new technologies such as chiplets, artificia Read more…

Baidu Exits Quantum, Closely Following Alibaba’s Earlier Move

January 5, 2024

Reuters reported this week that Baidu, China’s giant e-commerce and services provider, is exiting the quantum computing development arena. Reuters reported � Read more…

Leading Solution Providers

Contributors

Choosing the Right GPU for LLM Inference and Training

December 11, 2023

Accelerating the training and inference processes of deep learning models is crucial for unleashing their true potential and NVIDIA GPUs have emerged as a game- Read more…

Training of 1-Trillion Parameter Scientific AI Begins

November 13, 2023

A US national lab has started training a massive AI brain that could ultimately become the must-have computing resource for scientific researchers. Argonne N Read more…

Shutterstock 1179408610

Google Addresses the Mysteries of Its Hypercomputer 

December 28, 2023

When Google launched its Hypercomputer earlier this month (December 2023), the first reaction was, "Say what?" It turns out that the Hypercomputer is Google's t Read more…

Comparing NVIDIA A100 and NVIDIA L40S: Which GPU is Ideal for AI and Graphics-Intensive Workloads?

October 30, 2023

With long lead times for the NVIDIA H100 and A100 GPUs, many organizations are looking at the new NVIDIA L40S GPU, which it’s a new GPU optimized for AI and g Read more…

AMD MI3000A

How AMD May Get Across the CUDA Moat

October 5, 2023

When discussing GenAI, the term "GPU" almost always enters the conversation and the topic often moves toward performance and access. Interestingly, the word "GPU" is assumed to mean "Nvidia" products. (As an aside, the popular Nvidia hardware used in GenAI are not technically... Read more…

Shutterstock 1606064203

Meta’s Zuckerberg Puts Its AI Future in the Hands of 600,000 GPUs

January 25, 2024

In under two minutes, Meta's CEO, Mark Zuckerberg, laid out the company's AI plans, which included a plan to build an artificial intelligence system with the eq Read more…

Google Introduces ‘Hypercomputer’ to Its AI Infrastructure

December 11, 2023

Google ran out of monikers to describe its new AI system released on December 7. Supercomputer perhaps wasn't an apt description, so it settled on Hypercomputer Read more…

China Is All In on a RISC-V Future

January 8, 2024

The state of RISC-V in China was discussed in a recent report released by the Jamestown Foundation, a Washington, D.C.-based think tank. The report, entitled "E Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire