Enter Note Done
Go to previous page in this tab
Session
  • Presentation | GH31C: Early Warning Systems for Infectious Disease Based on Climate and Environmental Variability I Poster
  • Poster
  • Bookmark Icon
  • GH31C-0693: A machine learning approach to detect the seasonal risk of Vibrio spp.
  • Schedule
    Notes
  • Board 0693‚ Hall EFG (Poster Hall)
    NOLA CC
    Set Timezone

Generic 'disconnected' Message
Author(s):
Komalpreet Singh, University of Florida (First Author, Presenting Author)
Kyle Brumfield, University of Maryland
Anwar Huq, University of Maryland College Park
Rita Colwell, University of Maryland
Antarpreet Jutla, University of Florida











Vibriosis, a bacterial infection concerning public health, is clearly increasing over time and across locations. This study used computer modeling (Ecological Niche Modeling or ENM) to predict where and when Vibrio bacteria thrive. We tested different ways to make these models, and the Bootstrap method worked best. It was very accurate (scoring 0.97 out of 1.00) and gave consistent results. The most important environmental factors for Vibrio were phosphate, sea surface temperature (SST), and colored dissolved organic matter (CDOM). While CDOM didn't contribute as much to the model's initial training, it was still a critical influence on where Vibrio could live. Vibrio prefers: Warm water (around 25–28°C), Slightly alkaline water (pH around 8.2), Salty marine water (more than 35 PSU), Low phosphate levels, Moderate amounts of CDOM (which suggests clearer water). We also looked at how Vibrio habitats change with seasons. While overall hotspots stayed similar, a model focused on summer months (JJA model) showed much stronger and more concentrated areas of high risk. This means that focusing on specific seasons can help us predict and prevent Vibrio infections more precisely.














Scientific Discipline
Neighborhood
Type
Main Session
Discussion