Help working with large datasets

[GandarfTheWhite]

Hello, I just wanted to ask if anyone can give me tips or point me to any tutorials on how to build a graph and work with it when the dataset is extremely large (100gb).

I am working on node classification on the heterogenous TwiBot-22 dataset that comes in csv and json files.

I think I need to be using Database (SQLiteDatabase) and OnDiskDataset. Started by loading the edge list in with dask to process and trying to turn into edge_index and insert in SQLiteDatabase for each edge relation type (10). Please let me know if this is the right approach to start?

I am working on this for my undergrad computer science final project and am very new to GNNs so any advice will be much appreciated.
Thank you