I think you may be confusing the purpose of OrthoMCL, which is unrelated to functional annotation. The goal is to define orthogroups with which can try to work out orthology relationships, and then analyze patterns of evolution in a comparative way, usually in light of gene function or evolutionary rate. You did not say what type of data you have, but for many species, you can search for orthogroups by a term or function. If you included one of these species, or a species with annotations, you could carry over annotations from those reference species. If you are analyzing a group of non-model species, then you will have to generate these annotations yourself.
If you want to assign a putative function to your genes, run HMMER or InterProScan. The results can be mapped directly to GO terms, if that is what you are interested in.