Brainlit Project
NeuroData Design (2020-2021)
Advisors: Tommy Athey and Professor Joshua Vogelstein

Prerna's original cell detection scripts: https://github.com/prernasingh11/CLARITY

Files:

• CNN_soma_detection.ipynb: final project/script for testing Brainlit data on pre-trained model
• crossvalidation4.pt: pre-trained model
• paths.png: list of paths for training data if desired (did not use)
• requirements.txt: used to create python environment
• compareSubvols : used to run the model through a set of 12x12x12 subvolumes

Folders

• brainlit and cloud-volume: exisiting packages used to pull Brain1 data from the cloud
• output: NCRs and ROIs produced from CNN_soma_detection.ipynb
• TigerGPU:
  • slurm1.sh : runs wholeBrain.py on TigerGPU
  • testenv.yml : reproduce brainlit environment
  • wholeBrain.py : loop through entire brain, output all 12x12x12 subvolumes and their coordinates