BSI Seminar Social - Mosquitoes: olfactory-driven behavior and computer vision

Title: Mosquitoes: olfactory-driven behavior and computer vision

By: Evyatar Sar-Shalom

PhD student, Bohbot Lab, Department of Entomology

Hebrew University of Jerusalem

Abstract:

Mosquitoes use volatile cues to locate resources such as animal hosts and oviposition sites. Understanding how odorants shape these behaviors may help guide future vector-control strategies. In this seminar, I will discuss two compounds that reduce host-seeking behavior and a computer vision tool for measuring mosquito behavior. First, I will present evidence that indole reduces host-seeking behavior in landing and flight assays. I will then introduce YOLito, an open-source mosquito detector built on YOLO and trained on more than 38,000 images from six laboratories. I will show how it can quantify mosquito activity around oviposition sites. Finally, I will discuss borneol, a plant-derived repellent used for centuries, and evidence that the maxillary palp receptor OR49 contributes to female mosquitoes’ avoidance of borneol. Together, these studies connect mosquito olfaction with high-throughput behavioral measurement.

Short Bio:

Evyatar Sar-Shalom is a PhD student in the Bohbot lab in the Department of Entomology at the Hebrew University of Jerusalem. His research examines how odorants influence mosquito host-seeking behavior and how computer vision can improve behavioral measurement. He developed YOLito, an open-source mosquito detector created by fine-tuning YOLO, and uses behavioral assays and automated analysis to study mosquito responses to plant-derived compounds.

Students, please bring your OneCard.