17

This image is circulating on social media (e.g. Pinterest).

Eye color odds

It implies that if you have green eyes, then you can't possibly have one biological parent with blue eyes and the other with brown.

The image also says if you have brown eyes, your parents can't both have blue eyes.

Are these restrictions true?

Barry Harrison
  • 14,093
  • 4
  • 68
  • 101
not given
  • 171
  • 1
  • 3
  • 1
    So you go from "0%" to "can't possibly"? If the rate is 1 in 10000 we would still write that as 0%, but would not say "can't possibly". – GEdgar Nov 29 '15 at 01:11
  • 14
    @GEdgar: as this image has a `<1%` datapoint, your point is moot. – not given Nov 29 '15 at 01:37
  • 3
    The first problem I see with this chart is that it doesn't have hazel. – jwodder Nov 29 '15 at 17:19
  • 1
    Reverse searching the image could list some sites that explain about eye color inheritance. Also, related on Biology.SE: [How is eye color in humans inherited?](http://biology.stackexchange.com/questions/728/how-is-eye-color-in-humans-inherited) – Andrew T. Nov 30 '15 at 06:50
  • This subject fascinates me, as a 2-brown-eyed-parent family with a blue eyed and a green eyed child it shows how improbable that setup could be. I look forward to a good answer! – Jamiec Nov 30 '15 at 09:03
  • 3
    One issue with the chart: The chart makes assumptions about prevalence of genes in some general population, which the parents are assumed to be representative of, in order to guess likelihood of recessive genes in each parent. Those assumptions do not hold when sampling e.g. multiple children from same family. They do not hold if you already know other eye colours in your own family (e.g. grandparents). They also do not hold in many countries where different percentages of genes are prevalent. – Neil Slater Jan 30 '17 at 09:51
  • The apparent precision of the numbers is not necessarily representative; they all look like fractions of four - except the last one, which I can't explain -, so they are probably derived from principles, not from actual data. – tripleee Apr 11 '19 at 04:52
  • 1
    @tripleee I agree with your assessment. If I had to guess, the fractions of *8* are from a [dihybrid cross](https://en.wikipedia.org/wiki/Dihybrid_cross), which would be the result of assuming two genes are responsible for eye color. This is, however, [not true](https://www.nature.com/articles/jhg2010126). "**Although there are about 16 different genes responsible for eye color,** it is mostly attributed to two adjacent genes on chromosome 15, hect domain and RCC1-like domain-containing protein 2 (HERC2) and ocular albinism (that is, oculocutaneous albinism II (OCA2))." – Barry Harrison Apr 11 '19 at 05:48
  • I couldn't find any papers with percentages of offspring for different combinations of parents. – Barry Harrison Apr 11 '19 at 05:52

2 Answers2

10

According to the US National Institutes of Health (NIH) (emphasis added):

Researchers used to think that eye color was determined by a single gene and followed a simple inheritance pattern in which brown eyes were dominant to blue eyes. Under this model, it was believed that parents who both had blue eyes could not have a child with brown eyes. However, later studies showed that this model was too simplistic. Although it is uncommon, parents with blue eyes can have children with brown eyes. The inheritance of eye color is more complex than originally suspected because multiple genes are involved. While a child’s eye color can often be predicted by the eye colors of his or her parents and other relatives, genetic variations sometimes produce unexpected results.

In the question, you correctly noted that "The image also says if you have brown eyes, your parents can't both have blue eyes." This is contrary to the NIH's statement.

Thus, the figure is invalidated.

Are these restrictions true?

No, the second restriction is not true. The figure is inaccurate and at least one mistake exists.

Barry Harrison
  • 14,093
  • 4
  • 68
  • 101
-2

So naturally genealogy is a very complex field, very active research, and things are being refuted all the time. The problems seem to become more complex the more we know about it, so it might still be some time before you can "choose" the eye color of your offspring. A quick google search reveals on Wikipedia

"So far, as many as 15 genes have been associated with eye color inheritance. Some of the eye-color genes include OCA2 and HERC2.[9]"

Essentially the answer is that likelihood of inheritance of eye-color will need to take into account more factors. A nice little tool can be found here: which is NOT more accurate.

The real question is what accuracy will satisfy you to be correct? Ballpark figures: pretty much accurate. As accurately as we can predict: probably not

Underdetermined
  • 105
  • 2
  • 1
    Not to mention random mutations which make it impossible to make 100% accurate generic predictions. – Maurycy Nov 29 '15 at 16:39
  • 3
    [Welcome to Skeptics!](http://meta.skeptics.stackexchange.com/questions/1505/welcome-to-new-users) While science continues, the Wikipedia page you reference shows that we have a pretty strong research on eye-colour already. The tool you link to is pretty much a repeat of the claim (2-gene model). You don't show that the chart is even ballpark accurate, and you don't show the answers to the specific claims pulled out by the OP. – Oddthinking Nov 30 '15 at 01:29
  • Thank you for the comment oddthinking. I believed the claim "ballbark figures" would imply that the restrictions (0%) are not entirely accurate (even if the 2-gene model implies this). I assumed this question would not resonate that much with the community and wanted to encourage OP to do some of his own research, once he sees how easy it is to get an idea of the material using the internet. There is no need to do more to get to my conclusion, and perhaps a career of intensive study to gain a better understanding of predictive capabilities of genes. – Underdetermined Nov 30 '15 at 06:47
  • 4
    @Underdetermined : This community has certain norms. "Encourage OP to do some of his own research" is not one of them provided the OP brings a well-defined claim by a noteable source. The question resonates well with this community as you see by multiple people upvoting it. – Christian Nov 30 '15 at 11:33