Private Cloud at Leading US Genomics Institution

Case Study: Transformation of Genomics Workload Within Three Weeks

Unitas Global architected and implemented a private cloud solution with OpenStack to support genomic and cancer research.

The Challenge

Unlocking innovations in genetics and cancer research requires significant compute resources to analyze the massive amounts of genomic data and genome sequences made available by the Human Genome Project. But traditional methods of procuring and assigning these computing resources cannot keep up with the high—and fluid—demands of scientific queries. To eliminate bottlenecks and enable researchers to conduct data analysis more quickly and efficiently, the customer needed to implement a high- performance computing (HPC) solution that also allowed seamless access to both internal and external data resources.

The Solution

To support its HPC genomics and cancer research requirements, the customer decided to implement a private cloud platform using OpenStack. Architected to use existing hardware and software, the solution integrates several petabytes of genomic data stored in legacy systems into the customer’s private cloud while also enabling researchers to access important external data that can inform their research.

By moving from a traditional purchasing to a utility computing model, the customer can easily and quickly scale its computing resources, on demand, to support complicated scientific queries. Unitas also worked with the customer to implement a DevOps-style application lifecycle that can be used with private and public clouds to continuously update existing and rapidly release new applications used by the researchers.

Results

  • Within three weeks, the customer implemented a proof-of-concept private cloud and completed its first production genomics workloads three weeks later.
  • Automated processes and tools have enabled the customer to expand its cloud resources and deployment scripts, which can run workloads either internally, on its private cloud, or externally on a public cloud, have improved and sped up processes while also reducing the manual coding required by researchers.
  • The customer has also ensured that it continues to meet compliance requirements by integrating its private cloud into existing process controls.

Contact us for more information: