Warning messages: gene set enrichment analysis with clusterProfiler

Question

[adR]

Hello Dears, I was doing gene set enrichment analysis with a Bioconductor package clusterProfiler I am using window 10 and my R and Rstido are up-to-date but getting the following warnings while running KEGG gene set enrichment analysis. Here are Warning messages:

1: In preparePathwaysAndStats(pathways, stats, minSize, maxSize, gseaParam,  :
  There are ties in the preranked stats (0.01% of the list).
The order of those tied genes will be arbitrary, which may produce unexpected results.
2: In serialize(data, node$con) :
  'package:stats' may not be available when loading
3: In serialize(data, node$con) :
  'package:stats' may not be available when loading

I tried to google it to get an idea of what these warnings implying for but could not understand what is written out there. Anyone has an opinion; please share it with me.
Thanks!

Answer 1

[antonioggsousa]

Hi,

I'm not familiar with clusterProfiler. I know the package, but I did not work with it yet.

As far as I understand, gene set enrichment can be applied on rank-based methods. So, I believe this is the case. If it's using the p-values to rank your genes, let's say the gene with the lowest p-value will be the first, the second one with the lowest p-value the second, and so on, if you have genes with the same p-value, that means that you have ties in your rank list. This is two or more genes within the same rank. So, the message is that for these genes you'll get an arbitrary order.

I think you can check if you have the same p-values or not looking to the input data, that you're providing to the function that gives you that warning message.

At least this is my interpretation. You may want to post this question on bioconductor or github repo to get a more precise answer.

I hope this helps,

António

[adR]

Thank you so much! I will post it into Bioconductor for more tutorials.