Our seminar series runs on Thursdays from 10:20-11:15, in Jennings 360. The schedule for Spring 2020 is below. Talks are up to 45 minutes + questions. The schedule is below.
We particularly encourage graduate students, postdocs, and other trainees to present in our seminar!
If you would like to speak in the seminar, please get in touch.
Abstracts
Dr. Celeste Vallejo
Using Markov Mixture Models to Estimate Continuous-Time Rates from Discrete-Time Data with Application to Malaria Incidence and Recovery Rates
Abstract: Sampling a continuous time two-state stochastic process at discrete times and calculating transition probability matrices for each pair of consecutive observation times yields a time series of two-wave panel data; i.e. interval censoring. Estimating transition rates for the underlying continuous time process requires that we identify a time series of continuous time models whose transition probabilities at the observation times match those in the observed transition matrices. Empirical and theoretical literature over the past 43 years assesses whether or not the observed transition matrices are embeddable in the class of continuous time Markov chains, and if so, transition rates are calculated within that class of models. We show that non-Markov embeddable matrices are embeddable in the class of two-component mixtures of continuous time chains, but that a priori constraints are required to estimate transition rates in the resulting under-identified system. Depending on the imposed constraints, the rates in the mixture model may either be identified, or partially identified with resulting restricted ranges of non-uniqueness of transition rates.
We apply this methodology to estimate incidence and recovery rates from malaria infection in the Garki district of northern Nigeria in the 1970s. We are able to assess the impact of combinations of indoor residual spraying and distinct drug administration regimens where such evaluations had previously been regarded as not doable as a consequence of non-Markov embeddability of observed transition matrices.
Bio: Celeste Vallejo received her Ph.D. in Mathematics in 2018 from the University of Florida under James Keesling. She is a 2nd year post-doc at the MBI and is looking to make the transition from academia to industry. Her broad research focus is on building models to answer epidemiological questions. Since graduating, she has continued her graduate work in understanding population characteristics that contribute to the silent circulation of poliovirus in endemic regions. She has also started working with Markov mixture models and is using them to estimate malarial incidence and recovery rates. As a co-sponsored post-doc with Battelle she has been involved with creating a microsimulation model representative of the healthcare system in order to test interventions that promote patient care while also reducing costs to Medicare.
Daniel Brook
The relationship between Buprenorphine Waiver Certifications and Hepatitis C Virus Incidence in the United States, 2014-2017
Background: The United States (US) is currently experiencing an epidemic of hepatitis C virus (HCV) infection spread primarily through shared injection drug equipment by people with opioid use disorder. The Drug Addiction and Treatment Act of 2000 allows buprenorphine, an evidence-based medication for opioid use disorder, to be prescribed by physicians through waivers. The relationship between buprenorphine and HCV is unknown. Syringe service programs (SSPs) prevent the spread of HCV, but many communities do not have access to them. We aim to assess the relationship between buprenorphine availability and HCV incidence.
Methods: We obtained buprenorphine waiver certification data from the Substance Abuse and Mental Health Services Administration and calculated the patients per 100 people in each state that can be served via a waiver. We obtained acute HCV incidence data by state from the Centers for Disease Control and Prevention and SSP data from amFAR. We created three multivariable negative binomial models using a generalized estimating equation, selecting covariates through a directed acyclic graph. We also assessed for effect measure modification (EMM) by state SSP status.
Results: A higher than median buprenorphine availability in 2012 was associated with a higher acute HCV incidence in 2013-2017 (IRR: 1.66 95% CI: 0.91, 3.02). Higher buprenorphine availability per 100 persons in 2012 was associated with a higher acute HCV incidence in 2013-2017 (IRR: 2.90 95% CI: 1.05, 7.99). Higher buprenorphine availability per 100 persons in 2012-2016 was associated with a higher acute HCV incidence in 2013-2017 (IRR: 1.98 95% CI: 0.86, 4.52). We found no evidence of EMM on these relationships by SSP status.
Conclusion: Our ecological association model demonstrates that increasing buprenorphine availability was associated with a higher incidence rate of acute HCV in the US in 2013-2017. These results should be followed with a causal analysis of this relationship.
Bio: Dan Brook is a fifth-year MD/PhD student in the Ohio State Medical Scientist Training Program. Dan is currently a third-year PhD student in epidemiology in the College of Public Health with a graduate minor in translational data analytics. His advisor is Bill Miller, MD, PhD, MPH. Dan is interested in investigating the relationship between buprenorphine and incident hepatitis C virus infection among people with opioid use disorder. Dan’s clinical interests include preventive medicine and addiction medicine in primary care and infectious disease practices. His undergraduate degree is in Biomedical Science, also from the Ohio State University.
Dr. Boseung Choi
Statistical inference for epidemic models based on Bayesian approach
Abstract: I present new methods for Bayesian Markov Chain Monte Carlo-based in references in certain types of stochastic model for epidemic data. SIR (Susceptible-Infected-Removed) model is the classical method for modeling infectious disease spread. In this presentation, I introduce two methods for another modeling for epidemics. The first method utilizes the within-household network dynamics of a disease transmission. We apply it to analyze the occurrences of endemic diarrheal disease in Cameroon based on observational, cross-sectional data available from household health surveys. To analyze the data, we apply formalism of the dynamic SID (susceptible-infected-diseased) process that describes the disease steady-state while adjusting for the household age-structure and environment contamination, such as water contamination. The second method utilizes solutions to ordinary differential equations describing the large-population limits of Markovian stochastic epidemic models can be interpreted as survival or cumulative hazard functions when analyzing data on individuals sampled from the population. We refer to the individual-level survival and hazard functions derived from population-level equations as a survival dynamical system (SDS). we applied the SDS approach to data from a 2009 influenza A(H1N1) outbreak at Washington State University.
Bio: Dr. Choi is an Associate Professor in the Department of National Statistics at Korea University Sejong campus. Dr. Choi’s research includes the development of statistical estimation methods and their application to infectious disease data. His previous positions include an Assistant professorship in Statistics and Computer Science at Daegu University, and a post-doctoral position in the Department of Biostatistics and Epidemiology at the Medical College of Georgia. Dr. Choi was a visiting scholar at the Mathematical Biosciences Institute in 2014.
Dr. Guido España
Estimating the impact of dengue vaccination using agent-based models
Abstract: Dengue is the fastest spreading vector-borne disease worldwide with around 100 million cases occurring every year. In 2019, almost 3 million cases were reported in Latin America. Dengue interventions have been limited to vector control, but the first licensed dengue vaccine (CYD-TDV) offers a new tool to reduce the burden of the disease. Nonetheless, the implementation of the CYD-TDV vaccine has been limited given that the vaccine only protects people with previous exposure to dengue virus, whereas it increases the risk of severe disease upon infection on people without previous exposure. There are many challenges in the introduction of a dengue vaccine as a public health tool, such as the design and interpretation of clinical trials, as well as the projection of local and global impact. Mathematical and computational models can be useful to advance these interventions towards public health impact. We used an agent-based model of dengue virus transmission parameterized with empirical studies to support the design and interpretation of clinical trials of dengue vaccines, and to estimate the impact of routine vaccination in populations at risk of dengue. We simulated virtual clinical trials with our model to study different sources of bias in the estimates of vaccine efficacy against dengue. We found that heterogeneous transmission could bias these estimates up to 20% below the true efficacy, which could result in underestimating the public health benefits from dengue vaccination. Finally, we estimated the public health impact and cost-effectiveness of dengue vaccination with CYD-TDV in a local setting and at different transmission settings. Given the increased risk of severe dengue in people without previous exposure, we modeled a pre-vaccination screening strategy to vaccinate only those with confirmed previous exposure. The accuracy of this strategy depends on the sensitivity and specificity of the screening. Our results suggest that this vaccination strategy could be beneficial and cost-effective assuming a moderate level of transmission, moderate sensitivity and high specificity. This work highlights the importance of modeling to support public health decisions.
Bio: Guido España is a postdoctoral fellow in the Department of Biological Sciences at University of Notre Dame and a visiting postdoc at the Mathematical Biosciences Institute at the Ohio State University. Guido’s research is focused on supporting decision making with the use of mathematical and computational models to understand the dynamics of infectious diseases. In particular, his interests include the evaluation of vaccine impact and vector control in vector-borne diseases, such as dengue, chikungunya, and Zika.
Prof. Megan Meuti
How a better understanding of seasonal ecology of mosquitoes can lead to a reduction in West Nile virus
Abstract: The Northern house mosquito, Culex pipiens, transmit several pathogens including West Nile virus. Not surprisingly, the number of human and animal cases of closely tracks seasonal cycles in mosquito abundance, such that peak West Nile virus incidence occurs in the late summer and early fall. However, West Nile virus incidence is low in the late fall and winter, likely because during this time, females of Cx. pipiens enter an overwintering dormancy, or diapause, where they cease biting vertebrates and laying eggs, and instead divert all of their energetic resources towards survival. Even so, the low temperatures and prolonged absence of food takes its toll, and mosquito populations are at their lowest in the spring. Therefore, a better understanding of the when different populations of Cx. pipiens initiate and terminate diapause could lead to more targeted and efficient control of these pests. Preliminary data from our lab shows that mosquitoes from Lansing, MI enter diapause earlier in the year than mosquitoes from Lexington, KY. However, it is unclear how mosquito seasonal responses are shifting in the face of urbanization and climate change. Mosquitoes, like other animals, primarily rely on changes in daylength to accurately predict winter’s arrival. Evidence from our lab suggests that artificial light at night (ALAN) from human-illuminated structures and roads likely prevents mosquitoes from entering their overwintering dormancy, and that ALAN-exposed mosquitoes continue bite and lay eggs when exposed to diapause-inducing short days. Additionally, increased temperatures in cities due to the urban heat island effect as well as winter warming associated with climate change may also prevent mosquitoes from entering their overwintering dormancy. Taken together, these results suggest that urbanization and climate change might increase our exposure to mosquitoes and the pathogens they transmit, but may also provide opportunities to predict seasonal responses and better manage these disease vectors.
Bio: Dr. Megan Meuti is an Assistant Professor in the Entomology Department at OSU. She completed her PhD in the lab of Dr. David Denlinger at OSU in 2014. Her doctoral research demonstrated that Northern house mosquitoes, Culex pipiens, use their circadian clocks to regulate seasonal responses. After teaching at Kenyon College as a visiting professor, Megan joined the faculty at OSU in August of 2016. In addition to uncovering the molecular underpinnings of seasonal responses and mosquito reproductive physiology, her lab has begun to explore ecological aspects of overwintering dormancy in Cx. pipiens, how that might impact disease transmission, and how we might use this information to better manage and control mosquitoes.
Prof. Sarah Short
The mosquito microbiota and its implications for arbovirus transmission
Abstract: The mosquito microbiota is formed during mosquito development and persists into adulthood, when the mosquito transmits human pathogens. Previous work has shown that the microbiota plays a role in traits relevant to disease transmission, including susceptibility of the mosquito to viral infection. A clearer understanding of the factors that influence mosquito microbiota formation is therefore critical to our overall understanding of mosquito borne disease transmission and will help facilitate vector control techniques that involve manipulation or perturbation of the mosquito microbiota. We have shown that amino acid metabolism signaling in the adult mosquito digestive tract plays a role in determining the number of bacteria in the gut. We have also found that environmental factors such as nutrition during larval development can influence the composition of the microbiota and the overall size of the bacterial community well into adulthood. Additionally, our findings suggest that larval nutrition and the microbiota interact to impact mosquito longevity. I will discuss these findings and their general implications for mosquito borne disease transmission.
Bio: Dr. Short earned her Ph.D. in Genetics and Development at Cornell University in 2012 and continued her training as a Ruth L. Kirschstein postdoctoral fellow at Johns Hopkins University. Dr. Short joined the OSU Department of Entomology as an Assistant Professor in August, 2018. Her research is broadly focused on understanding variation in susceptibility to pathogen infection and transmission of infectious disease in insects. She primarily studies Aedes aegypti, the mosquito vector of dengue and Zika virus. She studies how insects interact with harmful and helpful microbes, the formation of microbial communities within insects, the impacts of commensal microbes on pathogen susceptibility and transmission, and the ecological and evolutionary forces shaping insect immune defense. As a vector biologist, she is also interested in finding ways to use this information to prevent the spread of mosquito-borne diseases.