A novel approach of age-grading of mosquitoes using SERS and machine learning models

Project Summary Mosquito-borne pathogens, including malaria, Zika, dengue, and chikungunya continue to be a major public health concern globally. As only older mosquitoes are infectious and represent a risk to human health, scientists have sought to age-grade mosquitoes based on this understanding; however, no reliable, cost effective and practical methods exist to age mosquitoes despite the tremendous epidemiological value of this approach. The overall objective of this R01 is to establish a novel approach to age-grade mosquitoes Aedes aegypti in the field. The approach we took is based on surface-enhanced Raman spectroscopy (SERS) to analyze the biomolecules from mosquito water extract that are bound with silver nanoparticles (AgNPs) and then the SERS spectra are used in modern machine learning models to age-grade the mosquitoes. Our central hypothesis is that AgNPs interact with specific biomolecules enabling SERS to generate unique and predictable spectral information for establishing modern machine learning models to determine the age of mosquitoes. Our prior work demonstrates the feasibility of SERS and Artificial Neuron Networks (ANNs) to determine the age of both lab (error <1 day) and field-collected (error < 2 days) mosquitoes Ae. aegypti. In the proposed work, we will establish robust lab and field-deployable protocols to produce reliable and repeatable SERS data of mosquito water extract. Then, we will manipulate the lab and field conditions to determine the impact of biotic (food and infection status) and abiotic (temperature) to SERS characteristics. Robust and accurate machine learning model based on modern ANNs and Domain Adoption (DA) strategies will be established and validated for age-grading mosquitoes in the field. In addition, we will explore the Multi-task Learning (MTL) strategies to simultaneously determine the age and infection status. Our long-term goal is to establish a rapid, cost-effective, and field-deployable system that enables real-time analysis and data sharing to facilitate epidemiological studies, risk assessment, vector control intervention monitoring and evaluation.