How DNAnexus Enables Human Longevity to Optimize Genomics Analysis and Interpretation for Healthy Individuals

When the first draft of the human genome was completed in the early 2000s, the possibility of using genomic data in routine healthcare seemed close to becoming reality. Now, more than two decades later, genomics has made inroads into healthcare — but it still has not made it into daily patient care. Scientists and clinicians at Human Longevity (HLI) are working to change that.

In a recent webinar, Wayne Delport, vice president of technology & bioinformatics at HLI, described the whole-genome sequencing product and clinical report that the company developed for physicians and patients in its Health Nucleus Clinic. He also talked about how the DNAnexus Platform is providing much-needed bioinformatics infrastructure to help the clinic ingest, analyze, and interpret the data. Watch the webinar on-demand. 

“Human Longevity was founded on the premise that the genome was going to revolutionize the practice of medicine and make a significant impact on the integration of genomics and medicine,” Delport said. “We’ve seen a lot of that in the oncology space, but we are not entirely convinced we’ve seen that in the primary care space.” 

To create a whole genome product that would benefit people coming in for a healthy screening, HLI launched the Health Nucleus, a data-driven personalized health platform that looks across all the data from a healthy individual. “It is intended to change the way healthcare is practiced,” Delport said.

Instead of the conventional approach to healthcare, which relies primarily on data from family history and physical examination, the Health Nucleus team collects data from whole-genome sequencing, imaging, and different kinds of biomarkers. Under this new model, physicians have access to AI-powered analytics to make sense of the patients’ data. 

A New Kind of Clinical Report

When HLI began developing its whole genome sequencing product, they quickly realized that there was no existing report structure suitable for describing the results in a manner that was engaging for both clients and physicians. So they decided to create one. 

The report needed to be simple enough for clients to understand while simultaneously providing useful information to clinicians including prognostic risk assessment data and pharmacogenomics information; however, it could not be used for diagnostic purposes.  

The current version of the HLI report has four sections; hereditary disease risk, pharmacogenomics findings, health predisposition, and nutrition and wellness traits. The  Human Longevity Clinical Laboratory (HLCL), a CAP/CLIA certified lab at HLI, issues these reports. 

The hereditary disease risk portion of the report provides data on pathogenic and likely pathogenic variants and disease risk. The pharmacogenomics section of the report captures information on star alleles and their functions as well as relevant drug response data. The next section presents data on polygenic risk. The final section of the report provides information on nutrition and wellness traits to help patients and clinicians discuss health habits, exercise needs, disease susceptibility, reactions to food, and more.

Powerful Data Analytics 

These reports are only possible thanks to the infrastructure for analyzing, interpreting, and storing the large quantities of data generated by whole genome sequencing assays. HLI currently partners with a CAP/CLIA validated reference laboratory for its whole genome sequencing needs, while it handles the analysis in house. HLCL is validated to assess single nucleotide variants and small insertions and deletions. 

To handle its bioinformatics needs, HLI previously relied on an in-house workflow that ran on Amazon Web Services. It included separate pipelines for secondary and tertiary analysis and interpretation. These were all internally developed, so the HLI team was responsible for the upkeep. The system also had a lot of components, including multiple PostgreSQL databases and application programming interfaces. 

“We were responsible for maintaining our workflows [and] maintaining our infrastructure,” Delport said. “We also had to develop our own privacy and security program.” They also had to optimize storage costs.

HLI recently began transitioning its analysis to the cloud-based DNAnexus Platform, including secondary analysis tools developed by Sentieon. When the transition is complete, they will have a much simpler technology stack that uses fewer components than the one they built in-house. 

“The great thing about this for us is that we have one SQL database with one API that we don’t have to develop. That is developed and maintained by DNAnexus,” Delport said. 

This also means that HLI scientists can focus on the important task of extracting valuable health information from the patients’ genomes and reporting their findings. 

“We don’t have to worry about any infrastructure management,” Deplort said. “The only thing people on our team have to worry about is building and deploying bioinformatics tools that are used in our reporting workflow.”

DNAnexus’ Platform meets all of the compliance standards required for secure storage and transfer of whole genome data, as well as those required for bioinformatics pipelines and data tracking according to the CAP Molecular Pathology checklist. “DNAnexus more than meets these requirements for us,” Delport said. ”We were comfortable with the security and traceability.”

Using DNAnexus is also more cost-effective per sample than HLI’s original infrastructure. Even at 10 percent of HLI’s current volume, doing the analysis on DNAnexus is 62 percent more cost-effective than using its internal infrastructure. And the cost savings increase as the volume of samples analyzed goes up.

HLI has begun piloting its whole genome reporting in some other clinics beyond the Health Nucleus. It is also developing a new product called the Genome Search Engine which focuses more on variant discovery and has a wider scope than the whole genome product. “We want to be an interface between a clinical care provider in the genomics space and genomic interpretation,” Delport said. 

Watch the webinar on demand.

Get in touch to learn more about how DNAnexus can support your analysis needs.

New Research Analysis Platform Enables UK Biobank To Exponentially Increase Size and Scale of World’s Most Comprehensive Biomedical Database

As the scope and data demands of the UK Biobank biomedical database gets bigger, we’re helping UK Biobank to meet its audacious ambitions. 

The latest improvement is a new cloud-based Research Analysis Platform (RAP), which will exponentially increase the scale and accessibility of the world’s most comprehensive biomedical database. Watch the short explainer video to learn more. 

Since 2006, UK Biobank has been on a mission to unleash the imaginations of the global research community to go further and improve human health. This project is another example of their dedication to democratizing access to this rich, multi-dimensional data resource.  

DNAnexus shares this vision.  We understand that delivering on UK Biobank’s promise of improving human health requires overcoming the challenges around scale and diversity of datasets, the need for integrated multimodal analytics and comprehensive analysis tools, stringent controls on data security and regulatory compliance, and the critical need for collaboration globally and across institutional firewalls.

That’s why we partnered with UK Biobank to build a scalable, secure, and collaborative Research Analysis Platform to accelerate the speed and scale of health-related research.

Enabled by DNAnexus technology and powered by Amazon Web Services (AWS), RAP provides approved researchers with the ability to access and analyze the 11 petabytes of data in the UK Biobank database from anywhere in the world. 

It also enables easier collaboration between researchers within the same team, or on the other side of the planet. Users can leverage a valuable library of tools for analyzing genomic and clinical data. 

And as the UK Biobank grows, the platform is designed to grow with it, a vital requirement as the resource is expected to surpass 40 petabytes of data by 2025. To put that number into perspective, it would take over a century of continuous viewing to consume 40 petabyte’s worth of high definition 4k movies!

We are proud to partner with UK Biobank on this landmark initiative that combines our leading biomedical informatics platform and insight tools with UK Biobank’s genomics and clinical datasets. This global collaboration brings us one step closer to accomplishing our vision of democratizing data access to drive innovations in research that profoundly impact patient lives.   

Learn more about the new UK Biobank Research Analysis Platform. If you aren’t an approved researcher, read more about UK Biobank and apply for access here.

The Hybrid Hackathons of the Future — now with Librarians!

Authors

Hannah Gunderman, Data, Gaming, and Popular Culture Librarian, Carnegie Mellon University Libraries

Ben Busby, Director, Solution Science and Principal Scientist, DNAnexus

Introduction

With the world still reckoning with the impacts of the COVID-19 pandemic, one thing that has remained constant is the need to change how people collaborate and communicate ideas, often shifting to remote and virtual formats.  The COVID-19 pandemic accelerated the rate at which hackathons are hosted in a virtual format. Remote hackathons have the potential to mirror the personally and professionally transforming experiences conveyed by in-person events to those who can not travel due to financial, physical, or environmental constraints. Remote hackathons allow the intellectual wealth of scientists in these countries to be applied to the important topics and goals of the hackathon, while supporting their health and safety through virtual participation. We hope that hackathons will retain a hybrid model to maximize the scientific contributions of both in-person and remote participants. 

Why are hackathons important?

Hackathons allow for concentrated, focused effort on a task or goal by bringing together scientific experts in a particular discipline, such as structural variants, or united by a common goal, such as ending neurofibromatosis.  Some hackathons solve thorny problems, make life easier for practitioners of specific disciplines, or push the boundaries of what a particular scientific field can do.  That said, hackathons not only produce content (usually software), but ideally also actively facilitate education and networking. Those who participate often have professionally transformative experiences that can lead to a wider scientific network, job opportunities, and increased confidence in their coding and research skills. 

Hackathons largely follow the model of “disruptive innovation” by serving as a prototyping layer across scientific organizations, producing new ideas and technologies that the community can then assess for value in their larger goals and initiatives. The prototypes that emerge either push the envelope of what is possible with biomedical informatics, or make day-to-day bioinformatics easier.  While the code isn’t necessarily persistent, these proof-of-concepts are intended for the community to build upon. Hackathons foster an environment with “buzz,” an economic geography concept referring to the serendipitous sharing of creative ideas that happens when people engage in face-to-face interactions. The last year has taught us that these benefits from hackathons are also afforded through hybrid or fully-remote formats, providing hope for a positive future of hybrid hackathons in scientific advancement and discovery. 

How do hackathons benefit the participants?

Not only do hackathons have an undeniable benefit to the broader scientific community, but, they also can provide transformative and impactful experiences for the participants themselves. These experiences largely revolve around the areas of confidence-building, educational development, and camaraderie. 

As described earlier, the “buzz” created in hackathon environments helps advance the sharing of creative and innovative ideas. Through this exchange of ideas, participants can advance their journeys in computational problem-solving and modern software development techniques. In the bioinformatics space, there are many beginner data scientists who are still learning foundational skills in computation and scientific collaboration. Hackathons, whether remote or in-person, offer a concentrated space for beginner data scientists to advance their skills in both of these areas alongside more established bioinformatics researchers. Not only does this afford educational benefits to these participants, but it can also increase their confidence as scientists who can contribute to important research endeavours. 

Finally, hackathons also create the opportunity for participants to forge close personal friendships and bonds, which can lead to long-term collaborations and network-building. 

Participants often find themselves in intensely challenging and time-limited environments as they race to accomplish the goals of the hackathon, and going through these transformative experiences together can lead to strong friendships and connections that span beyond the bounds of the hackathon itself. This is not limited to in-person hackathons, however: video-conferencing software such as Zoom and collaborative tools such as Slack allow participants to interact with each other and build both interpersonal and professional connections. 

A Retrospective Look Into CMU-DNAnexus Virtual “Genomic Data to the Clinic” Hackathon

The CMU-DNAnexus Virtual “Genomic Data to the Clinic” Hackathon (June 1st – June 4th 2021) was focused on bringing complex genomic data into the clinic.  Specifically, we focused on integrating Expressed Variants, Polygenic Risk Scores, Structural Variants and T-Cell Receptors into an Electronic Medical Record readable format using OMOP and worked on a clinically presentable interface. Remote support was offered by librarians from Carnegie Mellon University Libraries who have specialties in data management, bioinformatics, and information sciences. This support included collating important resources found by hackathon participants (such as tools, software, literature, etc.) into a single spreadsheet for easy access, reviewing the hackathon manuscript for syntax and readability, and preparing the manuscript for submission to BioHackrXiv. Communication platforms such as Zoom and Slack can offer ways to stay in touch and facilitate collaboration during a remote hackathon, but information can still get lost in translation in environments where we can’t see each other face-to-face. Librarians are trained in the information sciences and well-positioned to assist in keeping information organized and accessible during a remote or hybrid hackathon. 

Participants not only effectively used online collaboration tools to create innovative workflows and deliverables supporting the goals of the hackathon, but also used tools such as Slack to develop interpersonal friendships. Much of the same dynamic energy and “buzz” felt during an in-person hackathon was also felt in this virtual space and the experience has already led to some promising future collaborations and scientific endeavors, including an accepted proposal for a presentation at the 2021 annual meeting of the American Society of Human Genetics that will share the scientific findings from this hackathon. 

Upcoming hybrid hackatons

Although the pandemic is experiencing a long tail, we can still begin to envision what our post-pandemic future may look like, taking the lessons we have learned from navigating our remote environment for the past several months. One of the lessons we can bring into a post-pandemic future is that hackathons with a virtual option can help us create more equitable and diverse intellectual spaces for tackling the most pressing issues we face in bioinformatics. Moving forward, hackathons should take a hybrid model and allow for both in-person and remote participation, while allowing more team leads the sequestration they need to fully focus their energies on these efforts instead of juggling both work and the hackathon.

Further, leveraging the support of librarians in the hackathon space can lead to a more organized, cohesive, and collaborative experience for participants. This is particularly true for fully remote or hybrid hackathons, where clear communication channels are crucial for all participants. Librarians can help facilitate collaboration and coordination between remote and in-person participants, and help collate resources (such as tools, software, and literature) found during the course of the hackathon.  

We are excited to see what the future of hybrid hackathons holds for our field at large, and the scientific discoveries that will result from these events.  Below are some upcoming hackathons you can follow or get involved in!

Everything is bigger in Texas: Pan-Structural Variation hackathon in the Cloud! 

October 10-13, 2021, hosted by the Baylor College of Medicine

BioHackathon Europe

November 8-12, 2021, hosted by ELIXIR Europe

CMU-DNAnexus Hybrid “Genomic Data to the Clinic” Hackathon

March 9-11, 2022, hosted by CMU Libraries (stay tuned for more details!)

We also recommend keeping an eye on future events and initiatives hosted by the DEMON network, an international network for applying data science and AI to dementia!

Keep an eye on this link for more information about these and other events: https://biohackathons.github.io