Gene Expression

Ancestral quanta predicts phentoype?

ScienceDaily has a report on a presentation Mark Shriver gave at AAAS meething this year:

“We started with 22 landmarks on the faces that could be accurately located in all the images,” said Shriver.

These landmarks might be the tip of the nose, the tip of the chin, the outer corner of the eye or other repeatable locations. They then recorded the distances between all the points in all directions, so they had a distance map of each of the faces.

From their DNA profiles, Shriver could determine the admixture percentages of each individual, how much of their genetic make up came from each group. He could then compare the genetically determined admixture to the facial feature differences and determine the relative differences from the parental populations.

“This type of study, done on admixed populations shows that each person is a composite of their ancestors and that the range of facial features is a continuum,” says Shriver.
Shriver found that there was a very strong statistical correlation between the amounts of admixture and the facial traits.

Black Americans are on the order of 20-25% European in ancestry and 75-80% West African. But these proportions are average over the whole population. Individuals vary in blood quanta:


Common sense would tell us that the more ancestry one has from a population, the more likely it is that one should exhibit the traits associated with that population. But there is some variance, and as I have noted before individuals with the same blood quanta may look very different, and favor different ancestral populations. But the variance around expectation is contingent upon the genetic architecture of a trait. The greater the number of genes which effect variation on a trait, the closer the inheritance pattern is going to resemble “blending.” Since skin color variation is controlled a handful of large effect alleles segregating within the population it stands to reason that random chance is a powerful parameter when it comes to parent-child correlation of traits. More complex morphological characters might be controlled by so many genes that the likelihood that an individual will come to have an assemblage of alleles out of sync with its total genome profile could be sharply reduced (unless of course selection reshapes the frequencies of the alleles with the population).


  1. #1 Tom Bri
    February 14, 2009

    My kids are half Japanese, but hardly look it except for dark eyes and hair. People guess Spanish more often than Asian. But then, Mom’s family has big round type eyes, not very Asian looking. The noses are definitely east Asian, small and round. I picked out an non-Japanese-looking Japanese to marry, I guess.

    I am looking forward to cheap genetic typing, just to see what is back in both families.

  2. #2 razib
    February 15, 2009

    there’s a level of intersubjectivity to these perceptions. i’ve talked to asians who claim eurasian children invariably look more european, and europeans who claim they look more asian. hold individuals are “coded” depends

    1) on exogenous context (i.e., name)
    2) your own perspective

    e.g., some readers who were black were surprised that anyone could not agree that tiger woods obviously looked african american. but of course many asians see his asian features rather strongly.

    all that being said, we use only a few gross features to code for population. it seems plausible that this would not always reflect total ancestry.

    re: japanese, they’re about 25% ainu last i checked, on average. ainu are east asian, but never developed the “classical” features of east asians to the same extent as chinese and koreans (the japanese probably being a korean derived population).

  3. #3 dearieme
    February 15, 2009

    When I was at school we were told about the “Hairy Ainu”, so I guess that’s another trait atypical of East Asians.

  4. #4 Sandgroper
    February 15, 2009

    My half Chinese daughter attended Chinese schools all the way through her schooling from kindergarten to age 16. Almost all of her schoolmates were Chinese. She knew she must look half-Chinese, because I told her and she believed me (for example she has half epicanthic eyelid folds, and hair texture midway between East Asian and European), but she could never see it herself. She always thought she looked pure northern European.

    Then at age 17 she switched to an international school, where the students are a mix of Chinese, European, and about half of them are ‘mixed’ like she is. She says that after she had attended the new school for a few months, she could begin to see that she does look half-Chinese, and that what I had told her was true.

    I knew very well this happened with others, from Chinese people telling me that my daughter looks identical to me, and Europeans saying the same to my wife, but I was surprised that the exogenous context affected the way she perceived her own facial features.

    Now, she can’t discern the ancestry of the mixed kids at the new school – she has to ask me for advice. The kids categorize each other on nationality, not facial features.

    When we are among a strange crowd of Chinese, they all immediately recognize that she must be my daughter and pay no further attention. The Europeans are the ones wondering what I’m doing holding the hand of a beautiful young Asian woman. Well, it’s because I’m her Dad, and we’re good buddies.

New comments have been temporarily disabled. Please check back soon.