Skip to content
/ Cmito Public

Mitochondria detection in C. elegans EM images.

License

Notifications You must be signed in to change notification settings

jabae/Cmito

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

54 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Cmito: Mitochondria detection in C. elegans EM images

Cmito is a tool that enables dense reconstruction of mitochondria in C. elegans electron microscopy (EM) image stacks using convolutional neural network (CNN).

Setup

# Clone repository
git clone https://github.com/jabae/Cmito.git
cd Cmito

# Install requirements
pip install -r requirements

Cmito pipeline

Mitochondria prediction

python Cmito/detect_nucleus.py --image [test_data/test_image.png] --model [models/nucleus_model.chkpt] --output [output_path.tif]
  • Supports image files in tif, png, h5, cloudvolume images.

Mitochondria prediction models

  • models/dataset2_model25000.chkpt: Trained on EM images from Witvliet et al. (2021)1 Dataset 2.
  • models/dataset5_model104000.chkpt: Trained on EM images from Witvliet et al. (2021)1 Dataset 5.
  • models/dataset6_model166000.chkpt: Trained on EM images from Witvliet et al. (2021)1 Dataset 6.
  • models/dataset8_model122000.chkpt: Trained on EM images from Witvliet et al. (2021)1 Dataset 8.
  • models/dauer_model188000.chkpt: Trained on EM images from Yim et al. (2021)2 dauer dataset.

*Select appropriate model that your image is similar to the images each model has been trained on.

Mitochondria segmentation

python Cmito/segment_nucleus.py --mito [test_data/test_mito.tif] --output [output_path.tif]
  • Supports image files in tif, png, h5, cloudvolume images.

*[ ] are variables users can specify. Currently, the variables point to the test data.

Mitochondria reconstructions

Stage Link to reconstructions Source
L1 L1 reconstructions Witvliet et al. (2021)1 Dataset 2
L2 L2 reconstructions Witvliet et al. (2021)1 Dataset 5
L3 L3 reconstructions Witvliet et al. (2021)1 Dataset 6
Adult Adult reconstructions Witvliet et al. (2021)1 Dataset 8
Dauer Dauer reconstructions Yim et al. (2024)2 Dataset

Analysis

All codes for analysis and generating figures for Bae et al. (preprint) can be found in analysis.

Footnotes

  1. Witvliet, D., Mulcahy, B., Mitchell, J.K., Meirovitch, Y., Berger, D.R., Wu, Y., Liu, Y., Koh, W.X., Parvathala, R., Holmyard, D., et al. (2021). Connectomes across development reveal principles of brain maturation. Nature 596, 257–261. 2 3 4 5 6 7 8

  2. Yim, H., Choe, D.T., Bae, J.A., Choi, M.-K., Kang, H.-M., Nguyen, K.C.Q., Ahn, S., Bahn, S.-K., Yang, H., Hall, D.H., et al. (2024). Comparative connectomics of dauer reveals developmental plasticity. Nat. Commun. 15, 1546. 2