multiclass GSA: "contrast" for gene sets and saving output

[vmartyan]

Hi Mike and everyone,

I've done multiclass GSA on array data type for 4 classes. I have a couple of questions:

1. In terms of the results, I'm getting a list of positive gene sets only (please see the attachment). I realize that this is normal for multiclass response variable. I wonder though if (like "contrast" for each gene in multiclass gene analysis) it is somehow possible to get a "contrast" for each gene set as well. Because otherwise it's not intuitive how to determine which gene set is enriched in each one of 4 classes.

2. I'm unable to save output into a file because I'm getting an error about argument of length zero. Please see the attachment. I'm not getting this error for multiclass gene analysis done for this dataset and results are saved fine.

Dataset is also attached.

Thanks for your help,
Viktor

4groups_SAM_GSA.xlsx