• Microbiology meets machine learning
  • Fateme Yousef Saber,1,* Homa Keshavarz,2 Sara Azima Haghighi,3
    1. Department of Cell and Molecular Biology and Microbiology, Faculty of Biological Science and Technology, University of Isfahan, Isfahan, Iran
    2. Department of Cell and Molecular Biology and Microbiology, Faculty of Biological Science and Technology, University of Isfahan, Isfahan, Iran
    3. Department of Artificial Intelligence Engineering, Faculty of Computer Engineering, University of Isfahan, Isfahan, Iran


  • Introduction: Microbiology is defined as a scientific major focusing on small living creatures like bacteria, viruses, and fungi. Considering the fact that industrial manufacturing of essential products like enzymes, hormones, and medicines has attracted enormous attention, scientists are seeking the solution to develop sustainable approaches. Among all proposed green synthesis methods, microbial production seems to be the best option due to its fast, simple, and environmentally friendly production mechanisms.
  • Methods: The present article is a short review of how microbiology could benefit from machine learning algorithms. Various articles and books from Google Scholar and Pubmed database were used to gather this information using these keywords: microbiology, bioinformatics, and machine learning.
  • Results: Machine learning is a type of artificial intelligence in which computers will train rather than programmed to conduct specific tasks. Different algorithms counting Artificial neural network, Clustering, Decision tree, Gaussian process, K-nearest neighbors, Linear regression, Regularization, and Support vector machine/ regression have been widely used in different fields of microbiology: Analyzing Genomics, proteomics, and microarray data, evolution, and system biology. More specifically, machine learning has proved to be valuable in industrial strain development. Host strain selection, Metabolic pathway reconstruction, Tolerance enhancement, Metabolic flux optimization, Fermentation, and Downstream process are the well-studied subject in this area.
  • Conclusion: In line with industrial-scale production by microbes, it is essential to optimize crucial factors like temperature and pH. Applying computational modeling before conducting an experimental trial will provide a clear vision of how manipulating cellular and environmental factors would affect the process.
  • Keywords: microbiology, bioinformatics, machine learning