I am analyzing "proteomics" data generated by the Somascan assay . I have 44/1301 proteins that are complexes (such as IL23), so the protein has multiple UniProt or Entrez ids (one for each component of the complex).
I would like to perform gene (protein) set testing, but as far as I can tell, most gene sets do not accommodate complexes (instead they would contain the subunits i.e. IL23A and IL12B for the IL23 protein). This of course makes sense because they were developed using genes.
I am not sure if this situation also occurs in MS-based proteomics, but is there an established way of incorporating these proteins into gene/protein set testing? Are these simply dropped from the analysis or one of the subunits randomly chosen for standard gene sets, or are there proteins sets available that accommodate these complexes?
Note that I am not looking for set enrichment tests of protein complexes, but rather pathways or ontologies (i.e. KEGG, GO).
Thanks in advance!