Research
Research Thrusts
The following three primary research thrusts were identified as key gaps in PIPP capabilities and will be probed by this planning grant.
01
Thrust 1: Co-design of policy implementation, and risk analysis
- Early and sustained engagement with public health stakeholders for the co-design of pandemic prevention requirements
- Development and application of methods for decision making under deep uncertainty, value of information, and adaptive interventions
- Development of novel computational approaches exploiting advances in AI, data management, and HPC methods for creating integrated multi-fidelity, multi-method, and multi-spatiotemporal scale modeling analyses
Read More
Motivated by the need for new modeling approaches to incorporate the underlying and changing uncertainties into analysis outputs, we perform a Many-Objective Robust Decision Making (MORDM) pilot study of public health policies to address future pandemics. We focus on demonstrating MORDM to public health officials and on learning how the OSPREY platform could support widespread use of MORDM by researchers and public health stakeholders.
Co-Designing Capabilities for a Robust Pandemic Response: Stakeholder Engagement for Visioning, Backcasting, and Evaluating New Decision-Support Capabilities. Lima, Pedro Nascimento de, Abby Stevens, Raffaele Vardavas, Jonathan Ozik, and Robert J. Lempert. Santa Monica, CA: RAND Corporation, 2023. https://www.rand.org/pubs/working_papers/WRA3085-1.html.
02
Thrust 2: Robust data for modeling
- Development of novel real-time sensors and near real-time data streams from sensor-based air, wastewater, and human monitoring (including for novel pathogens)
- Integration of sensor, public health surveillance and clinical data through model-based, data assimilation approaches for combining data streams and epidemiological model forecasts
- Creation of large-scale open-science data storage and indexing capabilities for epidemiologic modelers
Read More
Computational epidemiological models have become trusted in silico laboratories, they need to be calibrated to empirical data, often from heterogeneous sources, while accounting for shifting realities. In this pilot study develop and apply data assimilation (DA) algorithms to sequentially estimate model parameters for multiple models, based on multiple empirical data streams, as data streams are updated.
Read More
As part of this pilot, we demonstrate a robust and non-invasive system of aerosol monitoring that is pivotal for epidemiology and viral surveillance by identifying the pre-indicators of pandemic. As a proof of concept for aerosol monitoring, the sensor sensitivity and selectivity is tested against COVID-19 (i.e., SARS-CoV-2 pseudoviral and viral particles) using their specific probes.
Read More
This pilot study seeks to build better data fusion and measurement strategies by leveraging novel deep learning and artificial intelligence approaches. In particular, we utilize data fusion techniques to develop real-time forecasting of FIB in the Chicago Area Waterway System (CAWS) using a diffusion convolution recurrent neural network model (DCRNN) framework in conjunction with training data collected via two state-of-the-art adaptive sampling sensor networks.
Read More
Prisons and jails were among the first congregate settings in which the emerging COVID-19 outbreak was recognized for its potential to spread rapidly. Incarcerated settings also impact the community, either through direct contact of inmates cycling through facilities or indirectly through contacts with staff. This pilot develops a dynamic repository of operational data of prisons and jails relevant to epidemiological surveillance and modeling.
Developing Distributed High-Performance Computing Capabilities of an Open Science Platform for Robust Epidemic Analysis. Collier, Nicholson, Justin M. Wozniak, Abby Stevens, Yadu Babuji, Mickaël Binois, Arindam Fadikar, Alexandra Würth, Kyle Chard, and Jonathan Ozik. 2023. In 2023 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW), 868–77. St. Petersburg, FL, USA: IEEE. https://doi.org/10.1109/IPDPSW59300.2023.00143.
NSF RESUME HPC Workshop: High-Performance Computing and Large-Scale Data Management in Service of Epidemiological Modeling. Stevens, Abby, Jonathan Ozik, Kyle Chard, Jaline Gerardin, and Justin M. Wozniak. 2023. arXiv. https://arxiv.org/abs/2308.04602.
Trajectory-Oriented Optimization of Stochastic Epidemiological Models. Fadikar, Arindam, Nicholson Collier, Abby Stevens, Jonathan Ozik, Mickaël Binois, and Kok Ben Toh. 2023.In 2023 Winter Simulation Conference (WSC), 1244–55. San Antonio, TX, USA: IEEE. https://doi.org/10.1109/WSC60868.2023.10408258.
Towards Improved Uncertainty Quantification of Stochastic Epidemic Models Using Sequential Monte Carlo. Fadikar, Arindam, Abby Stevens, Nicholson Collier, Kok Ben Toh, Olga Morozova, Anna Hotton, Jared Clark, David Higdon, and Jonathan Ozik. 2024. In 2024 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW), 843–52. San Francisco, CA, USA: IEEE. https://doi.org/10.1109/IPDPSW63119.2024.00151.
PSI/J: A Portable Interface for Submitting, Monitoring, and Managing Jobs. Hategan-Marandiuc, Mihael, Andre Merzky, Nicholson Collier, Ketan Maheshwari, Jonathan Ozik, Matteo Turilli, Andreas Wilke, et al. 2023. In 2023 IEEE 19th International Conference on E-Science (e-Science), 1–10. Limassol, Cyprus: IEEE. https://doi.org/10.1109/e-Science58273.2023.10254912.
A Portfolio Approach to Massively Parallel Bayesian Optimization. Binois, Mickael, Nicholson Collier, and Jonathan Ozik. 2021. arXiv. http://arxiv.org/abs/2110.09334 .
03
Thrust 3: Prediction of future pathogens
- Fundamental research in experimental and theoretical pathogen structure and evolution
- Scenario development for epidemiological modeling of emerging pathogens.
Read More
Coronaviruses have long been known to present a high pandemic risk, but they are not the only RNA virus of concern. Beyond SARS-CoV-2 and MERS-CoV, NIH has identified additional coronaviruses with no or limited treatments as having pandemic potential. This pilot aims to determine the sequence variation/mutations impact on enzyme specificity for SARS CoV-2 and an emerging pathogen (e.g., MERS-CoV, West-Nile, Chikungunya, Zika).
NSF RESUME Eco-Epi Workshop: One Health Surveillance and Predictive Intelligence for Eco-epidemiological Modeling. Stevens, Abby, Jonathan Ozik, Junhong Chen, Rachel Poretsky, and Arvind Ramanathan. 2023. OSF Preprints. August 30. doi:10.31219/osf.io/tazm2.