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

High Content Analysis (HCA) has become a cornerstone of cellular analysis within the drug discovery industry. To expand the capabilities of HCA, we have applied the same analysis methods, validated in numerous mammalian cell models, to microbiology methodology. Image acquisition and analysis of vari...

Full description

Saved in:
Bibliographic Details
Published in:PloS one 2019-09, Vol.14 (9), p.e0222528-e0222528
Main Authors: Petitte, Jennifer, Doherty, Michael, Ladd, Jacob, Marin, Cassandra L, Siles, Samuel, Michelou, Vanessa, Damon, Amanda, Quattrini Eckert, Erin, Huang, Xiang, Rice, John W
Format: Article
Language:English
Subjects:
Anti-Bacterial Agents - pharmacology
Antibiotics
Artificial intelligence
Automation
Bacillus megaterium - drug effects
Bacillus megaterium - ultrastructure
Bacteria - chemistry
Bacteria - drug effects
Bacteria - metabolism
Bacterial Proteins - analysis
Biology and Life Sciences
Bradyrhizobium - drug effects
Bradyrhizobium - growth & development
Bradyrhizobium - metabolism
Bradyrhizobium - ultrastructure
Cell culture
Colony Count, Microbial
Computer and Information Sciences
Content analysis
Drug discovery
Enumeration
Escherichia coli - drug effects
Escherichia coli - ultrastructure
Identification
Identification methods
Image acquisition
Image processing
Laboratories
Learning algorithms
Machine Learning
Mammals
Medicine and Health Sciences
Methods
Microbiological research
Microbiology
Microorganisms
Microscopy
Multilevel analysis
Phenotypes
Physical Sciences
Pseudomonas fluorescens - drug effects
Pseudomonas fluorescens - ultrastructure
Qualitative research
Research and Analysis Methods
Tetracyclines
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:High Content Analysis (HCA) has become a cornerstone of cellular analysis within the drug discovery industry. To expand the capabilities of HCA, we have applied the same analysis methods, validated in numerous mammalian cell models, to microbiology methodology. Image acquisition and analysis of various microbial samples, ranging from pure cultures to culture mixtures containing up to three different bacterial species, were quantified and identified using various machine learning processes. These HCA techniques allow for faster cell enumeration than standard agar-plating methods, identification of "viable but not plate culturable" microbe phenotype, classification of antibiotic treatment effects, and identification of individual microbial strains in mixed cultures. These methods greatly expand the utility of HCA methods and automate tedious and low-throughput standard microbiological methods.
ISSN:1932-6203
1932-6203
DOI:10.1371/journal.pone.0222528