Bug seed gene pipeline crash

[endast]

What

Fixes a problem in the seed gene discovery pipeline where a crash could occur if you had sample names that could not be cast to integers. #76 (comment)

To fix this we removed the unused --sample-file flag from the run-association cmd.

We also updated the seed genes config by adding a sample config for subsetting the samples and removed the batch_size from the example config (and also added a check to verify that you are not trying to pass the batch size via the config).

Also the test phenotypes data have been updated to include a non numeric sample name.

Testing

Running on the example data should be enough to test that it runs through.

You can add a batch size to the data_loader config to verify that the check works.

[HolEv]

LGTM!