Title:          Towards Non-Relational and Visual Analytics in VERSA
Speaker:     John Krulick
VERSA is python based framework for dataset analysis.  In this talk I will discuss the tool.  Then we will focus on extending the analytics to spatial and temporal data, tree/hierarchical data type, and visual analytics.

Title:          Adaptive Models of Zika Transmission
Speaker:     James Schlitt (with Alex Telionis, Daniel Chen, Dr. Bryan Lewis, and Dr. Stephen Eubank of the Virginia Tech NDSSL)
Following the distressing trend set by its viral cousins, Zika virus (ZIKV) has crossed the Atlantic and emerged as a threat in Americas. First isolated in 1947 in Uganda, ZIKV was not seen in the Western Hemisphere until February of 2014 when it began its march across the tropical regions of Chile and Brazil [1]. Phylogenetically similar to the Spondweni virus, ZIKV is a member of the large Flavivirus genus, which includes several other well-known human pathogens, most notably Dengue virus, Yellow Fever virus, West Nile virus, and Japanese encephalitis virus [2]. Illness caused by ZIKV is thought to be milder than those caused by the aforementioned relatives [3], but ZIKV is far from harmless. Implicated in severe intrauterine infections, ZIKV has been suggested as a cause for the increase in reports of microcephaly in Brazil [4]. Though this association has not yet been categorically established by the medical community, the rate of microcephaly in ZIKV affected regions has increased by twenty times since the arrival of the virus, and considering the high morbidity associated with the disease, this certainly merits further investigation. 
As one might expect given ZIKV’s relatives, the Asian Tiger Mosquito (Aedes albopictus) serves as a competent vector and is strongly implicated in the transmission of ZIKV [5]. Other Aedes species have been implicated in ZIKV transmission, notably A. hensilli in Micronesia [6], A. africanus in Uganda [7], and A. aegypti in Malaysia [8]. Unfortunately, these species are ubiquitous across tropical and subtropical Americas including parts of the southern United States. While this is unwelcome news for public health professionals, ZIKV’s reliance on the Aedes vectors allows us the opportunity to map the zoonotic transmission niche and begin modeling interventions. A better understanding of the factors driving the ongoing ZIKV epidemic will allow the evaluation of interventions, as well as the forecasting of future outbreaks. It may even be possible to uncover areas of systemic underreporting, which we would expect to find in more rural regions of northern Brazil.
The first step in our investigation will be the creation of geographic models of the zoonotic transmission niche of ZIKV using niche modeling software MaxEnt [9]. Inputs will include the mosquito presence maps from the VectorMap Project [10], population density from LandScan [11], climatological data, land cover data, and elevation data from WorldClim [12], as well as cases data from the Brazilian Ministry of Health. Once the niche model is complete we will turn to the agent-based network modeling software Epi-Fast [13] to model disease spread and intervention effectiveness in synthetic populations. Epi-Fast will be used to describe person to mosquito to person transmission as a result of mosquitoes biting susceptible individuals after biting others in the viremic phase of infection. The zoonotic transmission niche output will be used to set geographic rates of mosquito-borne transmission within Epi-Fast, which will then be calibrated to historic data.
The results of the simulation will be evaluated with regards to goodness of fit of the niche model, similarity of simulated epidemics to the calibration data, and outcomes of antiviral interventions. Additional subjects for exploration may include a niche model analysis of ZIKV comorbidities related to microcephaly, the impacts of projected climate change on population exposures, and the impacts of future encroachment as a result of projected industrial and agricultural development. The successful completion of this project will provide a novel model to explore ZIKV transmission, investigate potential surveillance gaps, map out areas of high risk, and plan interventions for future outbreaks.

Webex Meeting Num. : 644 714 417

Event Contact:
Ray Ren