29. The SpatioTemporal Epidemiological Modeler
What can public health officials and scientists possibly do to protect populations from emerging disease or to implement better response measures? The rise of global economies, the rapid international movement of people and products, and the increased reliance of developed countries on global trade, all greatly increase the potential threat from infectious and food borne disease. The near real-time monitoring made possible by an electronic health record will provide input for dynamic epidemiological models of diseases. The need to rapidly develop new models requires a community working to bring together data and models designed to assisting public health officials and scientists responsible for public health response and interventions. This community requires input from several scientific domains and should, therefore, be open and allow scientists to reuse, share, and extend, reference data and models, thus rapidly building upon each other’s work.
Available through the Eclipse Foundation, IBM Research created The Spatio-Temporal Epidemiological Modeler (STEM) as a tool designed to help scientists and public health officials create and use models of emerging infectious and food borne diseases. To date, the best numerical studies of epidemiology have been conducted in university research groups using custom scientific software. This approach has produced a foundation of knowledge that can be drawn upon as the field advances. Yet the traditional development of scientific software may be too slow to respond on the timescale relevant to an unexpected epidemic or pandemic. Clearly there is an urgent need for an additional approach that allows scientists to reuse data and models, to rapidly build upon the work of others, and to accelerate the development and testing of new simulation software.
Today the STEM community includes contributors from several university groups and organizations world wide. STEM uses mathematical models of diseases (based on differential equations) to simulate the evolution of disease in space and time. These models can aid in understanding and responding to the spread of such diseases. For example, STEM was used with the Ministry of Health in Mexico (GDF) during the H1N1 pandemic to rapidly determine the basic reproductive number of the virus. Israeli CDC has used STEM to study differences in transmission between type A and type B seasonal flu. New models have been developed for the effects of climate change on global malaria burden, and a recent paper to appear in the Journal of Theoretical Biology describes how STEM was used to compare three different models for Dengue Fever.
STEM comes pre-configured with a vast amount of reference or denominator data for the entire world.[5-8] By using and extending the data and models in STEM it is possible to rapidly prototype and test models for emerging threats to public health. STEM also provides tools to compare and validate models.
As an open source project, STEM is free and completely open to any scientist or researcher who chooses to build on and contribute to its growing library of models, computer code, and denominator data. The Eclipse public license grants all users “a non-exclusive, worldwide, royalty-free copyright license to reproduce, prepare derivative works of, publicly display, publicly perform, distribute and sublicense” their contributions. Platform independent, STEM is available in versions for Microsoft, Apple, and Linux operating systems. Under the Open Services Gateway Initiative (OSGi) standard as an Eclipse application, STEM offers “plug and play” software architecture. All of its main components – the core representational framework, graphical user interface, simulation engine, disease model computations, and various data sets – are partitioned into separate bundles or plug-ins. Each of these components can be independently developed, deployed, and used with declarative software extension points to build on existing models and create new ones, making STEM extensible, flexible, and re-usable
With this component software design, models and scenarios in STEM can be easily shared, extended, and built upon. Creating a framework for modeling infectious disease requires data and models across several disciplines. Recently STEM has been extended to support the study of vector borne, zoonotic, and food-borne diseases. Models of malaria and dengue fever are now available building upon global vector capacity (mosquito population) models. STEM contains ten years of historic climate data from NOAA and the NASA Earth Observatory for the study of how climate affects vector distribution. The STEM framework has recently been extended to support the study of disease transmission in livestock and how development of new microparasitic infections in livestock can transmit to the human population as part of the overall food chain graph. The latest release of STEM also provides a proof-of-concept model for the distribution and spread of Salmonella from animals to humans via consumption of contaminated meat.
Why it should be recognized:
Recognizing STEM will help encourage others to join the open source project and will make it more visible to public health organizations that might benefit from STEM. Modeling disease requires building models upon models. As the community grows, experts from different domains can contribute new denominator data and new models that will strengthen the STEM Framework.