Madhav Marathe

Director and Professor, Network Dynamics & Simulation Science Laboratory, Biocomplexity Institute of Virginia Tech

SuperNova Award Category:

  • Data to Decisions
The Company: 

At the Biocomplexity Institute of Virginia Tech, we strive to overcome today’s challenges in human health, security, and sustainability by focusing on information biology, using high-performance computers to translate complex biological data into actionable knowledge. We do this by adapting to the ever-changing research environment by investing in the best people and resources, creating transdisciplinary teams of experts that span the globe, and seizing opportunities to push our science forward by inventing cutting-edge tools and sharing them.

The Problem: 

Among various benefits of globalization ,one of the downside is ease of any virus  being spread to multiple places in a short span of time.Epidemiologist usually struggle to know impact, nature , behaviour and any predictive measures of new virus spread. Speed at which virus spread is much higher than time needed to make any preventive measures resulting in loss of humans life.It will help if Epidemiologist have access to a tool which can provide analytical and decision-support using disease models , synthetic populations ,interventions and predictive analysis  that provide unprecedented access and information to plan, respond, and control pandemics.

The Solution: 

We've developed novel analytical and decision-support tools that provide epidemiologists unprecedented access and information to plan, respond, and control pandemics. Our global synthetic information database is one of a kind and can be quickly reconfigured to provide situated decision-making capabilities. The technology has been developed and honed during its use by analysts for the H1N1, MERS, Ebola and the recent Zika outbreak. Our Comprehensive National Incident Management System (CNIMS) suite is mature enough that any analyst with minimal training can run complex computational experiments. The synthetic information system embedded in CNIMS is a novel approach to simulations. It allows researchers to study potential crises at the level of their primary, secondary, and even tertiary effects, providing analysts with an informed course of action. A distinguishing feature of the modelling environment is its ability to represent a broad range of interventions and behaviours.

The Results: 

This effort of modeling and simulation software development has positioned the institute to provide sustained, real-time epidemic decision support, using predictive models to inform effective relief efforts. Through years of supporting real world-driven demonstration studies, we’ve been able to continuously improve and refine our simulations, allowing us to provide support to several branches of the federal government. For agencies charged with coordinating the response to an international outbreak like Zika, the level of insight provided by CNIMS help public health resources be directed toward intervention efforts with the highest likelihood of success. Through this elegant suite of simulation systems, our researchers are able to provide us with the answers decision makers need to move forward. CNIMS also helps in representing individual and collective adaptive behaviors, giving the researchers an extremely detailed portrait of how individuals function in the face of disaster both on their own and as part of a larger society.

Metrics: 

Through the suite of simulation systems , our researchers were able to provide timely answers during outbreak of Ebola and Zika virus. We are now enhancing the model to work with third parties and enabling them to create different apps extending same ecosystem.

The Technology: 

- Supported platforms: Window, Unix- UI: HTML5,CSS3,Backbone.js, Javascript, JQuery, ArcGIS, D3 charts- WebServices : Java based Rest Services- Application Server/ Web Container: Jboss, Tomcat- Middleware: EJB, JMS, Drools- ORM: Hibernate- Cluster: Torque implementation of Portable Batch System (PBS)- Database : Oracle

 

Disruptive Factor: 

It has been interesting journey over years , started with a basic version of simulation system few years back , with limited computational power, single tenancy and limited analysis data to an enterprise level suite of applications.Initial version provided file downloads of analysis which are now real time graphical responses. Initial version of application was supporting single tenant however now they are multi-tenant supported with Real time data available . We now have high end cluster to perform complex simulations tasks in parallel and in a reliable way. It helps to serve researcher in an effective and efficient way.

Shining Moment: 

We do have a culture of organizing Codeathon time to time for improving existing set of features or evaluating a new technology adoption , working in a non formal way at times helps to take a nice break and often some valuable insights to the overall process. 

About Biocomplexity Institute of Virginia Tech

Research at the Biocomplexity Institute integrates disciplines from molecular science to policy analysis to address pressing challenges to human health, habitat, and well-being. We apply an information biology approach to predict, explain, and visualize the behavior of massively interacting systems.