demo data

[EdGreen21]

we should provide demo data - not sure how small a dataset makes sense. Possibly 100 cells, where one TCR is tumor reactive? This should be

1. Small to download (download should be possible so that users can see the proper data structure if they open the data in Seurat)
2. Insufficient to train a derivative model
3. Fast to run on the server (but we shouldn't pull cached results)

Kyle will look at this

[EdGreen21]

@KyleTCL could you make a small package of useful demo data - for example a fraction of the BT21 data?

[vvkkhjt]

+1

[EdGreen21]

@KyleTCL can you upload the public data from Caushi et al - this should be on our internal server anyway