Rank based - Average expression of a gene set.

Question

[rt.cgen]

I have a custom gene set that I am trying to compute average expression of across a patient list (sample data set below). Simple averaging of expression values across genes to get an average expression value for a gene set isn't the best way for obvious reasons.

I will greatly appreciate any help with a method or statistical analysis that can perform rank (?) based averaging for gene sets. I am well aware of GSEA, GSVA and the likes that give an enrichment score but they also require two condition sets and my data doesn't have that.

Sample data -

            Gene1   Gene2   Gene3 ... GeneX
Patient 1   10      15      1   ...
Patient 2   2       80      5   ...
Patient 3   0.5     150     9   ...
Patient 4   .       .       .   ...

While searching I found a similar question but it doesn't satisfactory answer my question.

edit - My goal would be to estimate a score to calculate distance between samples set (patients) based on an average expression of average expression of gene in the gene set.

Answer 1

[theobroma22]

I'm not quite sure what you are trying to do, but if it's for an analysis and not a method development I would suggest to use PCA or PLS.