I am working on a system that annotates variants in human DNA sequences. One of the pieces of information reported is the zygosity of the variant. I noticed that mitochondrial variants are also shown to be either homo- or heterozygous.

How does that make sense? Does it make sense? As far as I know, the mitochondrial DNA is just a single, circular DNA molecule. So I can understand considering all mitochondrial variants heterozygous, by definition. I could also accept that one could think of them as homozygous by definition. I just don't see how some variants can be homo- and others heterozygous. Am I missing something or is the annotation system I'm using not making sense?

  • $\begingroup$ Since all the mitochondria in a cell (I'm thinking mammalian, through it may apply to others) are derived from the mother, mtDNA should all be homozygous - in fact, I don't really understand how zygosity even makes sense when talking about mitochondria. I suppose it's possible that some mtDNA in a cell is damaged or mutated, likely through the action of ROS, so there theoretically could be two (or more) "alleles" in a cell if there is a mixed population of mitochondria. I don't know quite enough about mitochondrial genetics to make this an answer, though. Just my 2 cents. $\endgroup$
    – MattDMo
    Oct 21, 2016 at 13:55
  • $\begingroup$ @MattDMo ah, I hadn't considered mized populations, that makes sense. I guess it's theoretically possible to have ~50% of mitochondria having a given variant but I would expect that to be vanishingly unlikely. If there is a variant specific to a given mitochondrion, I don't really think there would be enough sequence data to be able to call it. I guess my main question is whether mtDNA is somehow doubled in a way that makes zygosity even remotely relevant. $\endgroup$
    – terdon
    Oct 21, 2016 at 14:07
  • $\begingroup$ as far as I know, the only doubling is mtDNA replication just before a mitochondria undergoes fission. You probably want to verify that, though. $\endgroup$
    – MattDMo
    Oct 21, 2016 at 14:38

1 Answer 1


You abosolutely can have mixed populations of mitochondria in a cell (it is in fact very common). Therefore when you sequence the mtDNA you will get a completely variable percentace of reads containing a variant from close to 0% to 100% (there is a lower limit based on the quality and read depth of the sequencing data).

You can't consider the variants to be heterozygous or homozygous as conventional zygosity doesnt apply. Heteroplasmic and homoplasmic are the terms which are used.

  • 2
    $\begingroup$ So how would I, or a variant caller, call one or the other? Some arbitrary threshold like ">70% reads carry the variant so it's homoplastic" and "~50% carry the variant so it's heteroplastic"? Is there some literature you could point me to? Your whole answer could really benefit from some references to back your claims (I know your claim is true, but you're not supporting it in you answer). Still, it sounds like it would make more sense to just look at allelic balance values and forget zygosity for mt vars. $\endgroup$
    – terdon
    Sep 9, 2019 at 14:49
  • $\begingroup$ Welcome to Biology.SE! Your answer would be even better if you included supporting references (primary literature is best). ——— You might also want to go through the tour and then consult the help pages for additional advice on [Answer effectively] on this site. Thanks! 😊 $\endgroup$
    – tyersome
    Sep 9, 2019 at 18:33

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .