Memory issue

[Cristinacondel1]

Hi!

I'm trying to run the function on 100756 cells for the GO genesets and I'm always running into the same memory issue:

In asMethod(object) :
sparse->dense coercion: allocating vector of size 17.8 GiB

I'm doing this on a cluster with quite a big capacity so I decided to fraction the cells to be able to run it. I was wondering if then the data from different runs would be comparable, because from what I understand from normal GSVA the results are only comparable intradataset.

Thank you!

[guokai8]

Hi @Cristinacondel1 ,
Yes. You are correct. It could be only comparable within the same data. I am working on the memory issues and will update the function in next version.
Best,
K

[guokai8]

Hi @Cristinacondel1 ,
Please try the new version by setting the batch parameter within the scgsva.
K