cas9-similarity-search

DNA similarity search using Cas9

A visual overview

You are currently in the Cas9 Similarity Search repository. The feature vectors we start with come from the preceding Open Images repository: https://github.com/uwmisl/primo-openimages

Connecting to jupyterlab in a docker

SSH Port Forwarding

On your local machine: Pick a port, P: anything unused; a large number. When connecting with ssh: ssh -L localhost:P:localhost:P brinstar.cs.washington.edu. This forwards port P on your local workstation to port P on brinstar.

Run docker

On brinstar: Launch the docker using the script in the cas9-similarity-search repository: sudo ./docker.sh -d <path to primo-openimages dir> -p P, where P is the port you forwarded when connecting via SSH. The path is /home/kstwrt/HDD/kendall/datasets/primo-openimages/

So an example command is: sudo ./docker.sh -d /home/kstwrt/HDD/kendall/datasets/primo-openimages/ -p 9000

Launch browser

Once you launch jupter notebook in the docker, you can connect in your browser. Copy the /?token=<secret token> bit from the jupyter output, and point your browser to: http://localhost:P/?token=<secret token>.

Launch shell in docker

sudo docker exec -it <docker id> /bin/bash You can find your docker id by doing sudo docker ps

In docker, find Jupyter token to connect to

jupyter notebook list

MISC Helpful Commands

For monitoring GPU: watch -n 1 nvidia-smi