I have a basic understanding of genetics, and I'm really puzzled by this. My grandma's blood type is A (I don't know if it's heterozygous or homozygous) and my grandpa is AB. Yet my mother is type O.

Grandma says that my mom is 100% my grandpa's child, so how did it happen? I have a few hypotheses regarding this. Either my grandma is lying, which I find to be really implausible, or there's some genetical anomalies either in my mom or my grandparents.

I suspect that grandma is heterozygous, having the IAi genes for blood type, and grandpa has a weird case of IAIBi alleles, coming either from a failed disjunction which caused both alleles for A and B to be present in one sperm cell, and the i allele to be present in the egg, resulting in the IAIBi genotype. Then, when grandpa had my mom, both i alleles got combined and out came my mom with O type.

My other hypothesis is that mom is heterozygous, either IAi or IBi, but the dominant allele underwent a deletion (or something else) and was inactivated, and as such only the recessive allele remained, giving her type O blood.

Could I get some insight on how this happened? I can't seem to find anything related to it on the internet.

  • 3
    $\begingroup$ Never discount the prospect that someone in the lab screwed up. $\endgroup$
    – user438383
    Commented Mar 18, 2021 at 15:52
  • 4
    $\begingroup$ Maybe have a look at the Bombay blood group. It is a pseudo 0 blood group. en.wikipedia.org/wiki/Hh_blood_group $\endgroup$
    – KaPy3141
    Commented Mar 18, 2021 at 16:15
  • 3
    $\begingroup$ The simplest probability is that is wasn't grandpa's sperm that did the job and that grandma either "jumped the fence" or was raped, and either way was afraid or ashamed to admit the truth. $\endgroup$ Commented Mar 18, 2021 at 23:39
  • 1
    $\begingroup$ One thing for you; the "failed disjunction" hypothesis is testable; if true a genetic test would report your mother to be the child of her father's non-extant brother. $\endgroup$
    – Joshua
    Commented Mar 19, 2021 at 14:25
  • 1
    $\begingroup$ As @KaPy3141 mentioned, other genes such as the Bombay blood type may suppress the A and B alleles (epistasis). There was once an episode of a soap opera that had this exact plot. $\endgroup$ Commented Mar 19, 2021 at 20:20

3 Answers 3


The ABO blood type is controlled by a single gene (the ABO gene) with three types of alleles inferred from classical genetics: i, IA, and IB. The IA allele gives type A, IB gives type B, and i gives type O. As both IA and IB are dominant over i, only ii people have type O blood. [1]

So from this standpoint, normal Mendelian genetics cannot explain the blood-group 0 offspring, given the AB parent.

However, besides errors in testing, the (pseudo?) 0-bloodgroup could be explained by the Bombay phenotype. I don't know the incidence rate in your region, so I can't comment on the likelihood.

To give theoretical background, this would be considered a pseudo 0-phenotype, as the individual genetically does not need to be ii. Instead, the individual lacks another effective enzyme that is required for the A or B alleles to be effective; the blood-type specific sugars (N-Acetylgalactosamin|Galactose) cannot be attached to the red blood-cell membrane, as the anchor of these sugars (called H-substance) is not present.

Finally, to address your hypothesis of a gene-duplication followed by a gene knock-out: There is no need for a preceding duplication. The more probable cause is a single point mutation (or multiple) near the active center of the inherited A/B enzyme, or a nucleotide insertion/deletion in its coding region, causing a frame-shift. A completely (not partially!) inactivated A/B enzyme would effectively represent a novel 0-genotype.

  • $\begingroup$ Your link says its incidence is typically 4 per million, up to 1 per 10,000 in select areas/populations. IMO, unless there's a family history of it, it still seems likely that somebody's lying $\endgroup$ Commented Mar 18, 2021 at 17:44
  • 3
    $\begingroup$ @Punintended that doesn't follow at all. 4 per million means that on average 4 people in a randomly selected sample of 1 million will have the genotype. Since we live in a world of several billion people, it is expected that some will have this genotype. Therefore, the fact that it is rare doesn't tell us anything about whether the grandmother is lying. Some people do have this, so why not the OP's mother? $\endgroup$
    – terdon
    Commented Mar 19, 2021 at 17:07
  • 1
    $\begingroup$ @Punintended yes, but the odds alone are meaningless without taking into account the size of the sample pool. Essentially, here we are asking "what are the chances that a random person selected out of a pool of ~8 billion is one of the 4 per million" and that is a very different question to "how likely is it that this person is one of the 4 in a million". $\endgroup$
    – terdon
    Commented Mar 19, 2021 at 17:23
  • 1
    $\begingroup$ @terdon If you have 100k people and an incidence of 4/million, stochasticity kicks in and the odds change. That's not the case for large sample sizes, like an n of 8 billion. 4/million * 8 billion is 32k. The odds this person is one of those 32k (given no other knowledge) is 32k/8b, or the incidence rate $\endgroup$ Commented Mar 19, 2021 at 17:37
  • 1
    $\begingroup$ @Punintended but we have other knowledge. So that’s irrelevant. $\endgroup$
    – Tim
    Commented Mar 19, 2021 at 17:55

Two highly unlikely, yet possible ways this could happen:

One extremely unlikely possibility is that grandmother was born O. Grandmother's current A typing could be the result of interplacental stem cell transfer from bearing an A child and the mother is an XX clone of the grandmother. How similar in appearance are grandmother and mother?

A less rare yet still doubly-unlikely scenario is that the grandmother and grandfather are both chimeras and at least some of the grandmother's eggs and at least some of the grandfather's sperm are O. We are learning that human chimeras are more common than we used to think. https://en.wikipedia.org/wiki/Chimera_%28genetics%29


Given that your grandmother is type A and your grandfather is AB, there is very small, minute chances that your mother is in fact the biological child of one or both of them. For your grandmother to say that she is 100% your grandfather's child, maybe he had her with someone other than your grandmother, and due to unfortunate events, that woman died giving birth or shortly thereafter. It may hurt your grandmother to admit that. Or, she could have the same views as many others, like myself, and recognizes that there is a BIG difference between being a father and raising a child, and just being a sperm donor and disappearing. There are many different situations that could be possible, that your grandparents may not be ready to admit for whatever reason, or it could very well be a simple mistake at the lab who did the testing. But does it honestly matter if she is the biological child of both grandparents, or whatever the truth may be? Will knowing the answer change anything?

  • 1
    $\begingroup$ Welcome to Biology.SE! On this site, answers are expected to be fact based (and are much more likely to receive a favorable response they include supporting references (primary literature is best)). Your answer is entirely opinion and thus is not appropriate here. ——— Please retake the tour and then consult the help center pages for additional advice on How to Answer effectively on this site and then delete or edit your answer accordingly. Thanks! $\endgroup$
    – tyersome
    Commented Mar 19, 2021 at 18:11
  • 3
    $\begingroup$ Your theory of "biological grandmother is not the grandmother" doesn't really solve anything; the grandfather is AB, so the child would be A, B or AB, even if the biological grandmother was O. You'd still need to explain away the grandfather's lack of O managing to contribute an O, making the possibility of an unknown biological grandmother non-explanatory. $\endgroup$ Commented Mar 19, 2021 at 19:34
  • $\begingroup$ Did not realize that I needed to specifically state that there is a possibility of the mother being adopted and never told, thus explaining why she has neither of their blood types. My apologies, I will be sure to be more specific next time. $\endgroup$
    – Jae Mac
    Commented Apr 9, 2021 at 9:02
  • $\begingroup$ Please edit your post to make it better or, if you don't feel that is worth your time. then please delete it (click on the "delete" next to "edit" under your post. Thanks! $\endgroup$
    – tyersome
    Commented Apr 26, 2021 at 22:08

You must log in to answer this question.

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