Genealogical vs. genetic ancestors

An interesting simulation by Luke Jostins:
The probability of having DNA from all of your genealogical ancestors at a particular generation becomes vanishingly small very rapidly; there is a 99.6% chance that you will have DNA from all of your 16 great-great grandparents, only a 54% of sharing DNA with all 32 of your G-G-G grandparents, and a 0.01% chance for your 64 G-G-G-G grandparents. You only have to go back 5 generations for genealogical relatives to start dropping off your DNA tree. [. . .]

The number of genetic ancestors starts off growing exponentially, but eventually flattens out to around 125 (at 10 generations, 120 of your 1024 genealogical ancestors are genetic ancestors).

3 comments:

Wanderer said...

Short Answer: No.

Long Answer:
He says we only receive "any DNA" from ~120 of our 1024 biological ancestors 10 generations ago? Sorry, but this contradicts what we know about genetics.

There is indeed a fall-off point for receiving any genetics at all from any given ancestral-slot. But it starts much further back than generation-6, as claimed there.

We can give a simple answer by assuming that each chromosome is passed on intact...
That isn't accurate.

DNA is not transmitted one chromosome at a time. It is passed on in small clumps of tiny nucleotides called kilobases. Every human being has something like 20,000 packs of kilobases of DNA. When new life is formed, it is a genetic lottery, and some of these clumps of kilobases are passed on, and others not, in the great genetic lottery. This way, eventually some kilobases disappear way down the road in any particular line-of-descent. *Any single ancestral slot* of JoeBlow beyond 14 generations back (when we all have 16,000 ancestral-slots) is statistically likely to not be a genetic ancestor, as any single ancestral-slot will likely have imparted 0 kilobases of DNA to JoeBlow. But, again, DNA doesn't get passed on one "chromosome" at a time. The blog's premise is wrong.

[See here: http://johnhawks.net/weblog/reviews/genetics/populations/inbreeding_ap_2006.html -- ctrl-f for the word 'billions'; that is the relevant section].


The simplest way to prove that this is not logical is the case of blackAmericans. Births of fully-phenotypically European babies to lightskinned black couples would be common if this idea that "only 43 ancestors impart any genetics to a person". That never happens. The racial-stock of the child always resembles that of the parents.

n/a said...

Wanderer,

The numbers I quote are based on a simulation using empirically-derived recombination rates (not the assumption of no recombination). The estimate is 120 genetic ancestors at 10 generations -- not 43. I have not examined Jostins' assumptions or execution in detail, but the results seem reasonable.

"DNA is not transmitted one chromosome at a time."

Yes, (after recombination) it is. Based on these observations (which Jostins uses), the average number of recombinations per chromosome per meiosis appears to be ~1.5 (higher in females; lower in males). The number of recombinations per chromosome is certainly not in the hundreds or thousands.

A kilobase is merely a unit of measure (one thousand bases).

Hawks is saying the same thing as Jostins, just using rough estimates rather than running a simulation and focusing 30-40 generations back rather than 1-10.

American blacks are around 20% white genetically. If, on paper, someone has 224 white ancestors and 800 black ancestors ten generations back, the probability that all their genetic ancestors will be white is (much) less than (224/1024)^120 = 6.21967087 × 10^-80. The chances remain infinitesimally small even if half the genealogical ancestors are white.

On the other hand, in the case of, say, an otherwise white person with one or a handful of Amerindian genealogical ancestors 10 generations back, there's a good chance he will have inherited no genetic material from them. (And, of course, many white trash "Cherokee princess" stories have no basis in reality to begin with.)

Rathssen said...

hey Wanderer,

If what the article claims is wrong, then explain this:

http://spittoon.23andme.com/?s=native+american,+5+generations


Published by one of the leading DNA testing companies out there, 23&Me, the content goes hand in hand with what was presented in the article here at www.blogger.com