VE Strat - V1: VE-Cadherin Adherens Junction Stratification by Deep Convolutional Neural Networks

VE-Stat

This is the data and belonging to the VE strat publication, a deep learning tool for VE-Cadherin cell border quantification and stratification

Training images for the model are included in the TRAINING folder. Examples of different stimuli are also included in the Examples folder

The pipeline we created is mostly geared towards the ImageXpress microscope filestructure, but can be adapted to other image sets as well Packages used by the scripts are listed in the dependencies.

Examples of how we implement the analysis is found in the "ImageXpress_pipeline.py" and the "liver analysis" files Images for these examples are located in the examples folder. Images need to be extracted from the tar archive first.

Examples of using it for a single image is found in the "single image analysis" file. The functions and classes are found in cell_border_identification_function_classes.py file.

The model itsself is saved as a keras model, and is the onyl keras file in here. The training procedure of the model is found in the training script, and the training data is found in the TRAINING folder.