The initial question was about understanding what is in the downstream of a gene in a eukaryotic organism. I understand that this region is located 3' of a gene, and therefore I would expect to find non-coding regions, but is there a difference between non-coding regions and an intergenic region?

A small background of what I am doing: I am studying the effect of retrotransposons in S. cerevisiae populations, and I already identified the potential impacts of these retrotransposons and these were the consequences:

Upstream gene variant   49
Downstream gene variant 42
Intergenic variant  4
Transcript ablation 3
Coding sequence variant 1
Feature elongation  1
3' UTR variant  1

So is it clever to say that downstream of a gene there are non-coding genes?

  • 1
    $\begingroup$ There may be non-coding regions (introns) within a single gene. $\endgroup$ – Vadim Jan 23 at 16:28
  • $\begingroup$ But up/downstream of a gene should not consider the introns as these are happening within the gene. $\endgroup$ – Silvia VC Jan 26 at 22:30

"Intergenic" is, well, an embarrassment, though it can be hard to avoid. Intergenic means, literally, between genes. Genes are, as you'd expect, genetically defined as regions of the chromosome with some genetic role. Between genes, we have, well, junk, spam, without meaning or significance, since otherwise, it would be a gene. In the 1990s everybody knew most of the genome was "junk DNA" and it was pretty straightforward. And the central dogma was DNA makes mRNA makes protein and if there was no mRNA making protein, or even no long protein, then they assumed there was no gene.

Well, the problem with using these terms nowadays is, first, it is very hard to show any given piece of DNA never has a role. After all, you can often remove a whole gene without visible effect on an organism, so how can some noncoding sequence 'between genes' be proved conclusively to be meaningless and of no relevance to either of its neighbors? And second, DNA definitely does have a role even if all it does is produce non-coding RNA, or act as a regulatory sequence that does. So for example, on OMIM you can look up many records for "intergenic", with results like H19, for which one term is "LONG INTERGENIC NONCODING RNA H19". Note also H19 is listed as the HGNC approved gene symbol, so it is an intergenic gene. Tricky, eh? (I have great respect for OMIM - this isn't their fault) Also, enhancers can be quite a long way from a gene, even past the end of another gene or affecting multiple genes, so even though a mutation in the enhancer might genetically act as an allele of a specific gene, it might be described in sequence terms as located in an intergenic region.

Bottom line: the intergenic regions you're reading about are going to be between things that are recognized as genes, and may or may not be known to have some effect on one or the other (or both...). They won't contain CDS, which is always considered to be a gene, nor 5'UTR or 3'UTR because those are transcribed before or after a CDS. Past that, how they are being distinguished from "upstream sequence" and "downstream sequence", that could be completely arbitrary. They might have gone to WormBase and pulled out curated gene records, and added some number of base pairs before the start site as promoter - but I don't know that. You'll just have to try to find the particular reference for whatever tool you are using.


An intergenic region is just one of many types of non-coding sequence.

Others include:

  • promoters
  • regulatory binding sites
  • terminators
  • introns
  • aptamers
  • ribosome entry sites

There's a lot of stuff that's part of genes besides the protein coding sequence.

A good way to get a picture of the sorts of things there are besides coding regions is to browse the Sequence Ontology. If you browse to coding sequence and start browsing parent and cousin terms, you'll find a vast number including all of the things that I mentioned above, as well as intergenic region, with definitions associated with ear.

  • $\begingroup$ correct me if I'm wrong but poly-A sequences are not coded into the DNA, they are the result of RNA processing $\endgroup$ – Hachiloni Jan 24 at 22:35
  • $\begingroup$ You're right; the poly-A attachment site gets marked, but it's a junction and not a region; I've dropped it from the list. $\endgroup$ – jakebeal Jan 24 at 23:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.