It is extremely important to develop novel analytical sensors for pathogen detection. Applications of such sensors range from ensuring environmental monitoring, to providing human health diagnostic measures, and to early warning systems for bioterrorism. Certain viruses and bacteria pathogens are of special concern for their ability to spread rapidly in human populations during a pandemic, and emerging variants of these pathogens are challenging to detect. This paper provides a reliable detection strategy to classify viruses from innocuous proteins. Our novel instrumentation is based on pyrolysis gas chromatography differential mobility spectrometry (PY/GC/DMS), and couples a genetic algorithm based machine learning method for pattern recognition with a microfabricated solid state biosensor. Together, this hardware and software combination allowed for an extremely high classification accuracy (94%) between the T4 bacteriophage from bovine serum albumin (BSA), an unrelated protein. This study suggests that this portable sensor can be used along with the sophisticated data mining method as a fast, reliable, and accurate tool for the recognition of viruses.
- differential mobility spectrometer (DMS)
- genetic algorithm
- neural networks
ASJC Scopus subject areas
- Electrical and Electronic Engineering