How meaningful is GO term enrichment of SNP outliers when only a small proportion could be annotated?

Question

[belinda-k]

Dear all, I was aiming to identify outlier loci under selection for populations of a South African bee species using RAD-Seq. Overall, I got 1,119 outliers using various outlier approaches. However, I was able to annotate only a small proportion (12 %, i.e. 136) in BLAST2GO and even less could be associated with GO terms, i.e. 59 loci. Thus, I was wondering how meaningful is it do GO term enrichment when most of the outliers are anyways not captured in the list of annonated locu?! Moreover, there is no reference genome existing for my species which complicates things even more. Please let me know what you think. Best wishes, Belinda.

[Gio12]

I don't know if this is relevant at this point in time, but it is to my knowledge that RAD-seq may not be appropriate for GO term analysis. RAD-seq is used to generate genomic data, not transcriptomic data. Genomic data does not necessarily reflect what genes an organism is expressing. In short, RAD-seq may not be very meaningful in regards to GO terms. Since this post is relatively old, I would be interested on what you ended up doing after your post. Thank you.