The mission of the Center for Computational Epidemiology and Response Analysis (CeCERA) is to bring together researchers and practitioners with diverse expertise to solve challenging problems in population health and epidemiology through the development and application of computational methods.

Ongoing research at the Computational Epidemiology Research Laboratory includes

RE-PLAN: Evidence-Based Response Planning

User-Centered, Data-Driven Response Planning with RE-PLAN

RE-PLAN uses population data at the individual or household level to make it easy to:

Click here to download a flyer which summarizes the features of RE-PLAN.

RE-PLAN was created at the Center for Computational Epidemiology and Response Analysis (CeCERA).

Support for the RE-PLAN project has been provided by the National Institutes of Health (1R01LM011647-01 and 1R15LM010804-01), the National Science Foundation (NSF 1514390), the Texas Department of State Health Services, and Tarrant County, TX.



Computational Methods for Quantifying Regional Ebola-Specific Resource Coverage

NSF 1514390

Data Sources (CSV):

CDC guidelines provide explicit information on the types and quantities of resources needed at healthcare facilities (HCFs) to address outbreaks of Ebola Virus Disease (EVD) or other High Consequence Infectious Diseases (HCID). Interim guidance for hospital preparedness specifies a multilevel approach that includes three levels of assessment and care HCFs that should be utilized in the event of an EVD outbreak: frontline healthcare facilities, Ebola assessment hospitals, and Ebola treatment centers. However, without an adequate understanding of the demand landscape, implementation of these guidelines are likely to prove ineffective. In this project, factors which would hinder the efficacy of CDC guidelines were identified, and methods to estimate resource needs and population demands for medical care during a complex EVD outbreak were developed.

To estimate the demand landscape, strategies were developed to quantify the needs of specific resources at each HCF by considering the type of each HCF, the geography of areas impacted, evolving disease dynamics, and populations surrounding each facility. A computational methodology was devised to consider the complex spatio-temporal dynamics of the disease as it spreads through the population. This included the development of a Decision Support System to allow decision makers to interact with computational models and define characteristics of individual HCFs using a graphical user interface.

Population estimates from the U.S. Census Bureau, service areas of health care utilization using Primary Care Service Area (PCSA) boundaries from HRSA’s National Center for Health Workforce Analysis, and spatial population partitioning algorithms were employed to quantify the resources needed at existing HCFs to satisfy population demand. Outbreak models were used to better understand how disease dynamics within a population affect the spatio-temporal fluctuations of healthcare resource needs. Differences in population distribution patterns exhibited by urban and rural areas, coupled with the uneven spatial configuration of the three types of facilities, introduce biases into the demand landscape which complicate the equitable distribution of Ebola-specific resources, thus leading to disparate levels of service among population subgroups. Findings of this project indicate that current CDC guidelines will lead to inequitable and inefficient levels of service and are likely to fail to adequately address the unfolding outbreak.


CDC Requirements
CDC Ebola Requirements
Texas Beds to Population Ratio for Ebola Assessment Hospitals
Texas Beds to Population Ratio for Ebola Assessment Hospitals

Modeling Disease Spread in Mass Gatherings

A mass gathering (MG) is an event, such as the Olympics, the World Cup, and the Hajj, which draws many thousands of individuals from around the world to a single location for some period of time. They can increase the geographic spread of infectious diseases as newly infected travelers return home. The goal of our research is to create a computational MG model to better understand the spread of infectious diseases and to use this understanding to develop new disease prevention strategies. Specifically, we are modeling the Hajj in which over two million Muslims from more than 189 countries travel to Makkah, Saudi Arabia annually.

Modeling Epidemic Progression in Non-Homogenous Populations

The progression of disease epidemics is affected by differences in individuals' physiology and social/behavioral characteristics. Further, the geographic distribution of the population plays an important role in shaping the progression of disease outbreaks. The goal of this research is to develop a computational model to quantify the effect these factors have on spatial and temporal variations of disease epidemics. Specifically, we are interested in studying models that consider the effects of the immune response and the different interaction parameters of individuals during an epidemic.

The Power of Disgust: Modeling the Effects of Disease Avoidant Behavior

Disease avoidant behavior can be observed across the animal kingdom, from tadpoles to humans. The fact that this behavior is so prevalent implies that it confers a significant evolutionary advantage. Current epidemic models, however, do not sufficiently consider this behavior. This research uses agent-based epidemic models to measure the effect of disease avoidant behavior on the spread of disease. Preliminary results suggest that the inclusion of disgust (as a disease avoidant behavior) leads to a reduction of the basic reproduction number, a slowing of the spread of infection, and a lower infection peak.


Preliminary results of SEIR Model incorporating avoidant behavior.

Using Bayesian Networks to Identify Causal Relationships to the Incidence of Chronic Disease

Due to the nature of epidemiology, data collected and studied are primarily observational rather than experimental. New methods for identifying causal relationships within large datasets must be developed to understand the prevalence of chronic diseases in populations. This understanding will lead to the creation of improved methods to control and mitigate the risk of chronic diseases. The goal of this research is to develop heuristics using Bayesian networks to identify causal relationships in epidemiological data.


Utilization of urban green spaces by bumble bees (Bombus spp.) in Denton County

Bumble bees (Bombus spp.) are essential pollinators of both cultivated and wild flowering plants. However, declines have been documented in many bumble bee species worldwide. Though urban encroachment is a likely driver of such losses, urban green spaces may serve as habitat islands for bumble bee populations. Our research aims to ascertain the importance of such areas for bumble bees in Denton County by incorporating field, GIS, molecular, and computational modeling methodologies. Preliminary results from 2013 samples confirm the presence of two declining bumble bee species in Denton County, Bombus fraternus (Smith, 1854) and Bombus pensylvanicus (DeGeer, 1773), and demonstrate that existing community gardens and urban wild spaces provide habitat for both species. Ongoing research involves a population genetics approach to understand the structure of local populations, as well as computational modeling to predict how these populations may change in the future.



Bombus fraternus at Clear Creek Natural Heritage Center, Summer 2013