Introduction/Overview

OpenNeuro is the single best source I know of for BIDS-compatible neuroimaging datasets.

Quality Control is not gauranteed!

Just because a dataset is published on OpenNeuro does not mean it is rigourously quality controlled. Check any datasets you download for gross errors before committing to using them for a project.

✅ On the HPC, start an interactive desktop, load the datalad module, and navigate to the dlad directory you created in the previous practice.

✅ Open tutorial_datalad_on_HPC_commands.md in VScode

In this lesson, you will complete the OpenNeuro section. You will learn to install a dataset, learn about its characteristics, get and drop specific binary image files, configure .gitconfig, and work with nested datasets.

Nesting

The use of nesting is critically important to keeping dataset handling fast:

If you store everything in one dataset, commits gets slower and slower!
It is also difficult to split the dataset into subdatasets after the fact.
So, it is better to generate too many nested datasets instead of too few.
Each subdataset has its own stand-alone history.
Note, all subdatasets are directories, but not all directories are subdatasets!

GIT submodules

Datalad nesting relies on git submodules. Learn more here: https://www.atlassian.com/git/tutorials/git-submodule#

Navigate to the correct directories before running a command!

Be careful to navigate up and down in the directory structure as indicated! I will not always explicitly print out the cd command. I also expect you to use other common commands without being explicitly told:
● ls to list files, and
● pwd to see where you are, and
● du -sh to determine how big a directory is.

🟢 DataLad 2: Pr (10)