developer: Nicholas Weir, Ph.D. Student, Denic Laboratory, Harvard University
emai l: nweir a.t fas do t harvard (dot) edu
This package was developed to identify cells, foci, and reticular structures from Z-stack fluorescence microscopy images (multipage TIFF format). It was developed during preparation of Weir et al. 2017. Please cite this manuscript if you use this package!
As with any python module, clone the repository into your PYTHONPATH.
This in-development module is intended for segmenting Saccharomyces cerevisiae cells (or other roughly elliptical cells) using the signal from a cytosolically-localized fluorescent protein in fluorescence microscopy z-stacks. It is not perfect, and segmented cells require pruning post-hoc to eliminate poorly segmented cells (~5-20% when imaged cells are present in a single layer, more otherwise).
Classes and methods for segmenting small foci using a Canny algorithm (or using more basic thresholding methods if desired - see docstrings). Has methods to save intermediate images/output images as well as outputting segmented objects for further analysis.
Based on PexSegment.py, the classes provided here perform two additional steps:
- merging contiguous objects to generate longer reticular structures
- emptying holes in donut-shaped objects to avoid erroneous segmentation
Classes and methods for merging objects acquired in different channels (e.g. assign peroxisomes to "parent" cells). This code is still in active development.
Classes and methods for identifying whole cells expressing a cytosolic marker fluorescent protein based on a Z-stack image. This was designed for roughly elliptical cells such as Saccharomyces cerevisiae. It does not do well when cells are not present in a single layer. See docstrings for additional details. This code is stil in active development.
For usage examples, see the parallel repository for Weir et al. 2017 figure generation scripts in The Denic Lab
Readme last updated 5.24.2017