Virginia Bioinformatics Institute (VBI), a research institute dedicated to the study of information biology, is using DataDirect Networks, Inc. (DDN) SFA storage engine-based GRIDScaler GPFS parallel file system to perform Ebola outbreak modeling for the US Department of Defense’s Defense Threat Reduction Agency (DTRA).
A team of 30 researchers and scientists on VBI’s Ebola rapid response team initially provided DTRA and West Africa’s Ministries of Health (MOH) with short-term forecasts on vaccine production and disease spread. As the number of Ebola cases climbed, VBI moved to agent-based computational modeling to provide more in-depth analysis, including regional travel and social interactions and their impact on the spread of the outbreak outside of the infected regions and into other parts of the world.
To assist with interventions and policy recommendations while helping to mitigate global epidemic spread of the virus, VBI requires a cluster of extremely performance compute and storage resources to create highly detailed models that provide vital insight.
Since 2011, VBI has relied on DDN storage to capture, store and distribute millions of simulations, which then are applied to dozens of computational models to fuel rapid response efforts around the globe.
With hundreds of unique combinations of parameters, many of VBI’s simulations can produce 100% data growth year-over-year, making storage scalability a key requirement. DDN technology enabled VBI to deploy a cyber-infrastructure to support the creation of a synthetic population totaling more than three-quarters of a billion people. The computational models include multiple variants such as detailed demographics, family structures, travel patterns and similar activities to help model what could happen as the outbreak spreads throughout West Africa.
VBI’s Network Dynamics and Simulation Science Laboratory (NDSSL) worked in partnership with the institute’s Advanced Computing and Informatics Laboratories (ACIL) to handle different workloads. In one instance, VBI was able to meet a stringent 48-hour deadline from the Department of Defense, when completing intensive modeling for recommended locations for emergency treatment units in Sierra Leone, Guinea and Liberia.
With DDN’s performance, versatile, scalable and reliable storage, VBI is positioned to respond when an epidemic or crisis arises, while providing much-needed insights based on detailed and complex computational models.
Scalable DDN storage meets data – and compute-intensive research demands
With DDN storage, VBI can scale its storage to address a mix of data- and compute-heavy computational requirements while keeping pace with rapid data growth.
DDN’s strength in bioinformatics is reinforced by integration with different data types and its ability to support a variety of modeling applications and data analysis tools, including EpiFast, a fast simulation for producing a broad class of interventions; EpiSimdemics, for yielding detailed and flexible, complex dynamic interventions with rich disease specifications; and Indemics, interactive, SQL-based queries of simulations and very fast specifications of novel interventions.
To increase collaboration with other scientists around the world, VBI is exploring the use of DDN’s WOS object storage platform while also considering the company’s integration with iRODS.
Kevin Shinpaugh, PhD, director of IT and HPC, Virginia Bioinformatics Institute, said: “With DDN, we’ve attained a fast, reliable parallel file system to handle all our different workloads. DDN was best suited for our mix of computational models while delivering both the capacity and performance demanded by multiple users who need to access different data at the same time.“