I'm having some trouble understanding the concept of ancestral allele. What exactly does it mean? What does it have to do with Identity by descent/state? What does it have to do(if anything) with SNP's? Also, how is the concept useful in genetics/evolution? I've tried reading these up on various websites, but I'm not satisfied with my understanding yet.



Let's start with the definition that has nothing to do with the rest of the question :). A Single-Nucleotide-Polymorphism (SNP) is a kind of genetic variation that you find in population. This genetic variation is defined as a variation caused by only one single nucleotide (as its name indicates it). For example if you have in the populations the two following variants at the same locus: AGCCGT and AGCTGT, then you have a SNP at the position of the C and T. This type of variation is very common. We estimate that there are about 10 millions SNPs in the human genome. We often use these SNPs as genetic markers.

Ancestral vs derived alleles

An ancestral allele or an ancestral trait (depending on whether you look at the phenotype or a t the genotype) is the trait/allele that was carried by the common ancestor of the taxon you consider. For example: as you may know the taxon Reptilia includes lizards, snakes, turtles, birds, mammals and some other lineages. In the taxon Reptilia, the scales is the ancestral state (still exist in lizards for example) and the hairs and feathers are derived states. Usually, the ancestral state is the one that carried by the most basal lineages.

Identity by state (IBS) and Identity by descent (IBD)

If two sequences are exactly identical, then we say that they are identical by state (IBS). Such identity can occur either through convergent evolution or through common ancestry.

To understand the concept of "identical by descent" (IBD) it is important to understand the concept of coalescence. A coalescent event is, when looking backward in time, an event where two sequences were actually the same sequence. Often IBD is defined relatively to a given threshold in time. If the coalescent event occurs after this event then the two sequences are not IBD. If the two sequences coalesce before this time threshold and that the two sequences are still IBS, then the two sequences are IBD

Bridging everything together

So, if you look at sequences in a given group of lineages and see that they are all exactly the same then they are identical by state. If you know that all these sequences come from a common ancestor and even they all coalesce to a single individual in the common ancestor, then all sequences are identical by descent. But some mutations may have occurred since the last common ancestor resulting in some SNPs, then all the sequences are not IBS anymore and therefore not IBD either. You can use such sequence variance in order to create a phylogeny for example.

Hope that helps.

  • 2
    $\begingroup$ According to you, Identity by descent does not imply identity by state. But according to wikipedia en.wikipedia.org/wiki/Identical_by_descent it's the other way round. I realize wikipedia could be wrong in a lot of cases, but could you read what it has to say? My initial confusion arose from this article. $\endgroup$ – user6417 Jun 17 '14 at 9:20
  • 1
    $\begingroup$ @Venkat You're right it sounds wiki and I don't totally agree! To my understanding, IBD does not imply IBS because of recombination (I ignored this point in my answer above for the moment) and mutations. I don't mean that IBS implies IBD either. IBS without IBD may occur due to convergent evolution. I haven't read the wiki article but indeed this citation from the wiki article "DNA segments that are IBD are IBS per definition" sounds wrong to me! I'd appreciate to have the opinion of some other people to confirm or invalidate my point of view. $\endgroup$ – Remi.b Jun 17 '14 at 10:57
  • 2
    $\begingroup$ I believe I have seen both ways of defining IBD before, but I guess it makes sense to require alleles to be indentical (ie IBS) in order to be "indentical by decent". This is also the definition John Wakeley uses in his Coalescent Theory Book. $\endgroup$ – Paul Staab Jun 17 '14 at 11:00
  • 1
    $\begingroup$ Is it really what John Wakeley's says in his book? I'll check this out tomorrow as soon as I'm back home to my library. If indeed some (or all) authors defines IBD such that it implies IBS, I will modify or remove my answer. $\endgroup$ – Remi.b Jun 17 '14 at 11:03
  • 2
    $\begingroup$ On page 11, it says "The concept of identity by decent [...] posits that two or more gene copies are descented from a common ancestor without mutation. That is, their identity is a direct reflection of their common ancestry, as opposed to the more general case of identity by state [...]." $\endgroup$ – Paul Staab Jun 17 '14 at 11:09

Your Answer

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