Pittsburgh Supercomputing Center 

Advancing the state-of-the-art in high-performance computing,
communications and data analytics.

The Public Health Applications Group

The Public Health Applications group develops novel computational tools, models, platforms and infrastructures to empower public health decision making. Formed in 2012 and led by Dr. Shawn T. Brown, team members have a wide range of expertise and experience in public health and in software development, and work closely with decision-makers around the world. Below are highlights of some of their major projects:

HERMES (Highly Extensible Resource for Modeling Event-driven Supply chains) Logistics Modeling Team

HERMES-generated visualization depicting maximum storage capacity utilization (red being the highest and blue being the lowest) at different storage locations in the country of Benin.

Essential items like food,  vaccines, medications, diagnostics and maternal products only do good if they reach the people who need them, and reach them on time.  The HERMES Logistics Modeling Team develops  computational tools to evaluate and improve the delivery and distribution of such important products. The team has developed the HERMES software package, which  can rapidly generate a simulation model to evaluate any supply chain and propose improvements to it. Led by Drs. Bruce Y. Lee (Johns Hopkins University) and Shawn T. Brown (PSC), since 2008 the team has provided decision support to improve vaccine delivery in Niger, Kenya, Benin, India, Thailand, Senegal and Vietnam,  working with a range of international partners including the WHO, UNICEF, PATH, LOGIVAC, Village Reach, and the Bill & Melinda Gates Foundation. For more information about the project, please visit our website at http://hermes.psc.edu.

Partners: International Vaccine Access Center (IVAC) at the Johns Hopkins Bloomberg School of Public Health and the University of Pittsburgh Department of Industrial Engineering

Funding: Bill & Melinda Gates Foundation

 

Apollo: A single, service-oriented point of access to public health modeling

The Apollo team is developing a common standard vocabulary and a corresponding service-oriented architecture to connect infectious disease models, surveillance data, visualization tools and user-focused decision support tools. The software and standard vocabulary are completely open-source and freely available to the community. The vocabulary allows users to specify detailed disease, intervention, and population data once, and then harness the power of several simulation models simultaneously for forecasting and preparedness planning. Such a platform allows developers to have a common point of reference for epidemic data, greatly simplifying the processes of accessing data and connecting to other relevant software, as well as removing ambiguity in the meaning and interpretation of data inputs and outputs to disease models. The team is led by Drs. Michael Wagner (U. Pitt), Bill Hogan (U. Arkansas) and Shawn T. Brown (PSC).  For more information or to download the software see https://code.google.com/p/apollo.

Partners: University of Pittsburgh  Department of Biomedical Informatics  and the University of Arkansas for Medical Services College of Medicine

Funding: The National Institutes of Health

 

CLARA: A spatially-explicit model of humans and vectors for Dengue transmission

Fighting vector-borne diseases such as dengue requires an understanding of the activities of both the vectors (mosquitos) and the hosts (humans), and how the two intersect.   Our group has developed a model that represents both the human and vector population as individual agents, allowing us to explore the synergistic effects of human and vector interventions to determine optimal strategies for controlling these diseases. Dr. Nathan Stone (PSC) leads this modeling effort . The ComputationaL ARthropod Agents (CLARA) model merges realistic social  and demographic human  population data and the complete vector life-cycle to provide a novel platform for evaluating location- and time- dependent combinations of host and vector oriented interventions.

Partners: University of Pittsburgh School of Public Health and Johns Hopkins University

Funding: The National Institutes of Health and the Bill & Melinda Gates Foundation

 

RHEA: The Regional Healthcare Ecosystem Analyst

Our team has developed RHEA (Regional Healthcare Ecosystem Analyst), a software platform that can generate a simulation model of the healthcare services in any region. For example, we used RHEA to develop a model of the community and all the hospitals, long term acute care facilities, nursing homes, and patient flow amongst these, in Orange County, California. RHEA could help to evaluate and plan healthcare services and the introduction of new strategies and technologies in any country or region. The RHEA Team is led by Drs. Bruce Y. Lee (Johns Hopkins University), Susan Huang (UC, Irvine),  Kim F. Wong (University of Pittsburgh) and Shawn T. Brown (PSC).

Partners: Johns Hopkins Bloomberg School of Public Health, University of California at Irvine and Center for Simulation and Modeling, University of Pittsburgh

Funding: The National Institutes of Health

 

GAIA: A Geospatial Public Health Visualization Service

GAIA based visualization of an influenza epidemic in Allegheny County, PAGAIA is a web-based  visualization package which creates meaningful, interactive geospatial representations of public health and epidemiological data. The GAIA software translates this complicated data into a geographic representation and allows users to drill down to any level of the data that they want. GAIA provides visualizations that are more than just qualitative in nature. It also enables exploration of the quantitative data in an intuitive and unique manner, making it a truly useful tool beyond the “creating pretty picture” aspects of many visualization packages. GAIA is open-source software and is deployed at the PSC as an open web-service so that anyone who would like to use the software for visualization can do so. It is the visualization platform for the Apollo system as well as many more public health tools. More information is available at http://gaia.psc.edu.

Partners: University of Pittsburgh School of Public Health and University of Pittsburgh Department of Biomedical Informatics

Funding: The National Institutes of Health

 

MISSION: MIDAS Software Sharing and Information Outreach Network

MIDAS is a consortium of infectious disease transmission modelers. The MISSION group was formed to organize MIDAS software developers around the world to share software, provide best-practices in development and validation, and gather requirements for user focused tools, creating a true collaborative software development community.  Led by Dr. Shawn T. Brown (PSC), the group also meets regularly to engage in “hackathons” where their talents are focused on specific problems in public health research such as the use of Twitter feeds to gather outbreak data. The MISSION group keeps the community regularly updated on new software tools and developments through its Facebook page at https://www.facebook.com/MidasMission and conducts workshops to educate researchers and policy makers in the breadth of tools available through the network. MISSION has elevated the dedicated work of the software developers to its proper place in the research community and raised awareness of the power of simulation and modeling for action in public health.

Partners: RTI International, Inc., NIGMS MIDAS Research Network

Funding: The National Institutes of Health

 

PGRR: Pittsburgh Genome Research Repository

Personalized medicine—tailoring diagnosis and treatment of individuals to their own physiology, medical history, and genetic characteristics—has the potential of revolutionizing patient care and saving lives from diseases such as cancer. A major barrier in moving this field forward is the lack of highly available secure data streams for genomic data. The PGRR is a partnership between the Department of Biomedical Informatics and the Center for Simulation and Modeling at the University of Pittsburgh and the Pittsburgh Supercomputing Center to provide such a resource. By combining the domain expertise in state-of-the art genomics analysis, a multi-Petabyte storage infrastructure called the “Data Supercell” (http://psc.edu/index.php/computing-resources/data-supercell), and a high performance graph analytic appliance named “Sherlock” (http://psc.edu/index.php/computing-resources/sherlock), the Repository will give high-speed access to 100s of terabytes of data from The Cancer Genome Atlas (TCGA). The project is led by Drs Rebecca Crowley (U Pitt), Michael Barmada (U Pitt) and Shawn T. Brown (PSC) . The resulting web-based portal will catalyze life-saving research around the country.

Partners: University of Pittsburgh Department of Biomedical Informatics, University of Pittsburgh Center for Simulation and Modeling

Funding: University of Pittsburgh

 

Previous Work

VecNet Cyberinfrastructure for Decision Support

The currently available tools for fighting malaria cannot eliminate it. New strategies require an understanding of how interventions (things like bed nets, insecticides or vaccines) and other factors (the amount of rainfall, the temperature) affect the transmission of the disease. As part of the VecNet team, the Public Health Applications staff  helped to develop an innovative, versatile platform which integrates large entomological, epidemiological, demographic, intervention and climate data sets with complex disease transmission models to produce  computational tools tailored to the needs of specific communities that are creating new malaria-fighting strategies. The Public Health Applications group led the development of the Computational Intervention Portfolio Evaluator (CIFER) with Dr. Bruce Lee (Johns Hopkins University) and of  the Product Impact Evaluator tool (PIE) with Dr. Robert Farlow (R. Farlow Consulting, LLC.). CIFER links intervention portfolio management and evaluation tools with disease modeling, allowing  policy makers and disease control officials to determine optimal strategies for disease prevention and control. PIE enables product intervention developers to explore target product profiles through simulation modeling. The cyberinfrastructure is currently being applied to malaria control and prevention, but in the future it can also be applied to other public health related decision making.  For more information and to demo the current version of the cyberinfrastructure, visit http://vecnet.org.

Partners: The VecNet Consortium

Funding: Bill & Melinda Gates Foundation