Back

mAIcrobe: an open-source framework for high-throughput bacterial image analysis

Brito, A. D.; Alwardt, D.; Mariz, B. d. P.; Filipe, S. R.; Pinho, M. G.; Saraiva, B. M.; Henriques, R.

2025-11-12 microbiology
10.1101/2025.10.21.683709 bioRxiv
Show abstract

Quantitative analysis in bacterial microscopy is often hindered by diverse cell morphologies, population heterogeneity, and the requirement for specialised computational expertise. To address these challenges, mAIcrobe is introduced as an opensource framework that broadens access to advanced bacterial image analysis by integrating a suite of deep learning models. mAIcrobe incorporates multiple segmentation algorithms, including StarDist, CellPose, and U-Net, alongside comprehensive morphological profiling and an adaptable neural network classifier, all within the napari ecosystem. This unified platform enables the analysis of a wide range of bacterial species, from spherical Staphylococcus aureus to rod-shaped Escherichia coli, across various microscopy modalities within a single environment. The biological utility of mAIcrobe is demonstrated through its application to antibiotic phenotyping in E. coli and the identification of cell cycle defects in S. aureus DnaA mutants. The modular design, supported by Jupyter notebooks, facilitates custom model development and extends AI-driven image analysis capabilities to the broader microbiology community. Building upon the foundation established by eHooke, mAIcrobe represents a substantial advancement in automated and reproducible bacterial microscopy.

Matching journals

The top 11 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.