Use of high-content analysis and machine learning to characterize complex microbial samples via morphological...

Area of Science:

• Microbiology
• Cellular Biology
• Machine Learning

Background:

• High Content Analysis (HCA) is a key technique in drug discovery for cellular analysis.
• Current HCA methods are primarily validated in mammalian cell models.
• Microbiological methods often rely on low-throughput, labor-intensive techniques.

Purpose of the Study:

• To adapt and apply High Content Analysis (HCA) methodologies to microbiological samples.
• To leverage machine learning for microbial image analysis.
• To enhance the speed and scope of microbial quantification and identification.

Main Methods:

• Image acquisition and analysis of microbial samples, including pure and mixed cultures.
• Application of machine learning algorithms for quantification and identification.
• Validation of HCA techniques on diverse microbial populations.

Main Results:

• HCA enabled faster cell enumeration compared to traditional agar-plating methods.
• Successfully identified "viable but not plate culturable" microbial phenotypes.
• Classified antibiotic treatment effects and identified individual strains in mixed cultures.

Conclusions:

• HCA methods are successfully extended to microbiology, offering significant advantages.
• These automated techniques improve throughput and expand the utility of cellular analysis in microbial studies.
• This approach accelerates microbial analysis for applications like drug discovery.