What is the default/best practice to map prob IDs to gene IDs

Question

[Nemo]

Hi,

I have some datasets that have prob IDs and before performing gene set enrichment analysis, I need to map prob IDs to Entrez IDs. The problem is that after mapping, there are multiple prob Ids (multiple rows) per gene. I would like to know what are the best practices for picking one row per gene?

Since I wanna use gsea to analyze my data, I tried to find something about this in their resources, however, I wasnt successful to find what strategy they recommend for mapping from multiple prob IDs to a single gene ID.

If you have any idea, resources, or information about recommended solutions, please guide me.

Answer 1

[barslmn]

You can use biomart for this. There are we prob sets in filters or you could use your prob positions.

[Nemo]

Thanks @barslmn, I know how to find matched gene with each prob. But the problem is in the input expression matrix after mapping to genes, there can be more than 1 row per gene (lets say three probes per gene), and each probe has its own measurements in input expression matrix. How I can combine/pick one single measurement for each gene.

[barslmn]

You should check these out:

https://support.bioconductor.org/p/92128/