Gene filtering for FGSEA

[SirKuikka]

Hi,

I have scRNA-seq data and I am using log2 fold changes as the gene-level stats.

How should I filter the genes prior to FGSEA?

I have all 36 000 genes in a vector like this:

> head(fcs1)
MIR1302-2HG     FAM138A       OR4F5  AL627309.1  AL627309.3  AL627309.2 
 0.00000000  0.00000000  0.00000000  0.05287563  0.00000000  0.00000000 
> length(fcs1)
[1] 36601
> length(unique(fcs1))
[1] 14121

And the analysis warns of 3.8% of ties in the gene stats

> fgsea_results1 <- RunFGSEA(fcs1)
  |===========================================================================| 100%

Warning message:
In preparePathwaysAndStats(pathways, stats, minSize, maxSize, gseaParam,  :
  There are ties in the preranked stats (3.87% of the list).
The order of those tied genes will be arbitrary, which may produce unexpected results.
> 

Are those ties causing any issues? How would you filter the genes prior to GSEA?