Code and documentation to analyze the Allen Institute's Cell Type dataset
We could spend all day going through the cells in the website database to find some interesting comparisons. But, it’s much quicker to do these comparisons by accessing the entire database directly, through the Allen Software Development Kit (SDK).
If you’re at UC San Diego, you should run this code on the UCSD Data Hub (Option 2 below). Otherwise, you can run it on Binder, or locally on your computer.
This is the easiest option (if you’re not at UCSD), though the code itself will run more slowly. Simply click the button below to launch a dashboard containing both notebooks. Once you’re done with the Introductory notebook, open CellTypesNotebook.
While either on campus or connected via VPN, go to http://datahub.ucsd.edu and log in.
You’ll land on a page where you’re asked to choose a container to spawn. Choose the container with your course name.
On the Jupyter home page, go to New > Terminal (upper right corner).
Type (all in one line): git clone http://www.github.com/ajuavinett/CellTypesLesson.git
to clone the necessary codebase into your folder.
Close the Terminal window. On the Jupyter dashboard, you should now see a folder CellTypesLesson. Click into it, and open “Introduction to Jupyter Notebooks” From here, you can follow the directions in the Jupyter Notebook.
Once you’re done with the introductory notebook, open the CellTypesLesson notebook.

If you’d like to access the SDK locally, follow these steps. Note: If you’re in BIPN145, you should do this lab through the Data Hub (Option 1 above).
cd Documentsmkdir PythonNotebooksgit clone http://www.github.com/ajuavinett/CellTypesLesson.git
to clone the lab codebase into your folder.pip install --user allenSDK
Note: If you receive an error, try pip install allenSDKjupyter notebook