Here we report an intelligent workflow for sub-nanoscale 3D reconstruction of intact synapses from serial section electron tomography. Our workflow contains rapid ROI locating, automatic alignment, volume reconstruction, and semi-automatic synapse reconstruction. Multiple intact synapses in wild-type rat were reconstructed at a resolution of 0.664 nm/voxel to demonstrate the effectiveness of our workflow.
- Dependencies
- Instructions for Use
- Examples and Comparison Results
- Serial section ET data
- Contributing
Our workflow is mainly based on python and matlab. The required libraries are as follows.
python3.8, numpy, scipy, tqdm, tifffile, skimage, matplotlib, opencv-python, opencv-contrib-python, torch, torchvision, dgl, connected-components-3d, networkx, Pillow, connectomics
If you don't have some of these libraries, you can install them using pip or another package manager.
For the installation of connectomics, see https://github.com/zudi-lin/pytorch_connectomics.
In addition, we also use the pre training network of Super-SloMo. For details of Super SloMo, see https://github.com/avinashpaliwal/Super-SloMo.
If you want to use our workflow to reconstruct your target structure from serial section electron tomography, you only need to execute the files in order.
For a detailed description of our workflow, please refer to the paper "An intelligent workflow for sub-nanoscale 3D reconstruction of intact synapses from serial section electron tomography".
Here is a example of intact synapse reconstructed by our workflow at a resolution of 0.664 nm/voxel. And we compare it with the results of IMOD, TrackEM and Irtool.
We published our data on ScienceDB. The link is https://www.scidb.cn/s/eEVbqm
Please refer to the paper "An intelligent workflow for sub-nanoscale 3D reconstruction of intact synapses from serial section electron tomography".