Out of Africa: geographyic parsimony errors

Keith Robison robison at ribo.harvard.edu
Sun Nov 17 21:47:50 EST 1991

arlin at ac.dal.ca writes:

>In article <robison.690235642 at ribo>, robison at ribo.harvard.edu (Keith Robison) writes:
>[. . . stuff deleted]
>> 	I believe that the molecular evidence for an African origin of humans
>> is rather certain.
>> 1. mtDNA sequence analysis (see ref. below) generates a tree in which one must
>>    climb eight nodes from the root before finding a branch which contains
>>    non-African sequences on both sides of the branch (the tree was rooted
>>    with a chimpanzee sequence).
>> 2. One can climb several more nodes from the point described above before
>>    reaching a branch with large numbers of non-African sequences on both
>>    sides of the branch.
>> 3. The three closest ancestors of humans, pygmy chimps, common chimps, and 
>>    gorillas, all live in Africa (pygmy and common chimps diverged relatively
>>    soon after the chimp--human--gorilla split).
[stuff deleted] 
>> Keith Robison

>With regard to 1 & 2: 

>What's all this business about counting branches?  Ever seen it before?
>Ever seen anyone do a "hypergeometric test"?  What is it, anyway?  Don't
>molecular systematists usually do a "bootstrap" to test the significance
>of a particular branching order, instead of inventing their own method?

	Not branches, nodes.  What I meant by this is that if a node 
separates the set into {African} and {Non-African} subsets, then that node
must be accounted for by a to-Africa migration event under a non-African 
origin hypothesis.  Since we can count up eight nodes, the non-African
hypothesis is immediately handicapped with eight migration events. 
	Vigilant et al used a hypergeometric test NOT to test the validity
of the tree they obtained, but instead to test the validity of idea that
a tree rooted in Africa implies an African origin.  You can find a
hypergeometric test described in any freshman statistics book (I may have
the advantage here as I am closest to this ancestral state).  Here is the 
definition from mine (1, pp.242-243):

	1. The experiment consists of randomly drawing n elements without
	   replacement from a set of N elements, r of which are S's (for 
	   Sucess) and (N-r) of which are F's (for Failure).

	2. The hypergeometric random variable x is the number of S's in 
	   the draw of n elements.
An obvious example of a hypergeometric situation is a card deck and drawing
poker examples.  In the case of "Eve", Vigilant et al were estimating the
probability of "drawing" so many deep-rooting African mtDNA types randomly.

>Why didn't Vigilant, et al (1991) just do a bootstrap?  (hint: they wanted
>to support their conclusion, not weaken it).  And even if they could find
>a way to get a significant rooting to give an African clade on one side of
>the tree, does this really lead one to *conclude* an African origin?

	Yes, they did not report bootstrap results and they should have 
(they claimed that their data set was too large (note 36) -- I don't 
have any experience with bootstrapping so I can't judge this statement).  
However, they do report that their procedure generated 100 trees, and that 
these trees differed only in terminal branch placement.  

>With regard to  #3:

>Chimps and gorillas are *not* "ancestors" of humans-- they are RELATIVES.

Ooops! Yes, I know better and I goofed.

>Far from being another one of my characteristically vacuous semantic 
>arguments, this distinction is very important to evolutionary inference.
>The facts that chimps and gorillas are i) the closest living relatives of
>humans; and ii) limited to Africa, would only be relevant to this discussion
>if we had no paleontological evidence on the ANCESTORS of humans.  Human
>ancestors that are NOT chimp/gorilla ancestors include australopithecines 
>and the _homo_ species _habilis_ and _erectus_.  Pre-sapiens _homo_ fossils
>have been found in Indonesia (ever heard of "Java man"?), China (ever heard
>of "Peking man"?), Vietnam, Greece, France and Africa.  Since not all of 
>these places are in Africa, we may rightfully say that the pre-human 
>hominids, including the most likely candidate for a human ancestor 
>(homo erectus), lived in Africa, Asia and Europe. 

>Arlin Stoltzfus

>Department of biochemistry
>Dalhousie University
>Halifax, Nova Scotia 

	How can you tell these are species in the genetic sense?  
That they are indeed ancestors and not relatives? Treeing based on physical 
characteristics can be very tricky (for example, read (2) or another reference 
on panda phylogeny by O'Brien).  I believe that Jane Goodall's data on chimp 
skeletons also shows that you must be very careful in trying to classify 
species on the basis of old bones.  We can say that hominids lived in 
Africa, Asia, and Europe, and that these hominids pre-dated the emergence
of modern man, but IMHO we cannot use the bones to say where pre-human
hominids were distributed.

In another posting arlin at ac.dal.ca writes:

>In article <robison.690333981 at ribo>, robison at ribo.harvard.edu (Keith Robison) writes:
>> lamoran at gpu.utcs.utoronto.ca (L.A. Moran) critiques my migration tracing:

>...even if we could treat population migrations as
>character-state changes in lineages, we would not come up with the
>numbers that Keith Robison derives. A parsimony analysis is carried
>out by using allowable character state transitions to explain a
>distribution of observed character states within a pre-specified
>phylogeny. In the analysis that Keith Robison is presenting, the
>observed character states are geographic locations (Africa or
>not-Africa) of the bearers of mtDNA types, and the phylogeny is the
>supposed phylogeny of the mtDNA types.  The observed character states
>for extant lineages are given on p. 1505 of the September 27 issue of
>Science (Vigilant, et al., 1991), along with the tree.  The arguments
>below will not make sense unless you are looking at the diagram on p.
>1505.  The question that Keith is trying to answer is this: how many
>character state changes (Africa-to-not-Africa or not-Africa-to-Africa)
>does it take to explain the observed character states if A) the
>ancestral character state was *Africa* or B) *not Africa*.  His
>answers are A) 12 and B) 28.

>The first answer is correct but the second is incorrect.  Looking at
>the tree on p. 1505, it is easy to see that if the ancestral state was
>*Africa*, then changes to *not-Africa* are needed for 1) type 23; 2)
>type 28; 3) the ancestor of 49 & 50; 4) type 58; 5) the ancestor of
>types 74 through 135.  In addition, many descendants of the ancestor
>of types 74-135 went back to Africa, so we must postulate
>back-migrations for 6) type 76; 7) the ancestor of types 77 & 78; 8)
>type 83; 9) type 100; 10) type 103; 11) the ancestor of types 105-107;
>and 12) type 127.  Keith seems to have correctly identified all of
>these events (at least he came up with the right total of migrations
>and back-migrations).

>If the ancestral state was *not-Africa*, then changes to *Africa* must
>be postulated for 1) the ancestor of types 1-9; and 2) the ancestor of
>types 10-135.  Back migrations will then have to be postulated for 3)
>type 23; 4) type 28; 5) the ancestor of 49 & 50; 6) type 58; 7) the
>ancestor of types 74 through 135. In addition, many descendants of the
>ancestor of types 74-135 went back to Africa, so we must postulate
>migrations for 8) type 76; 9) the ancestor of types 77 & 78; 10) type
>83; 11) type 100; 12) type 103; 13) the ancestor of types 105-107; and
>14) type 127.  [these numbers should look familiar-- see previous
>paragraph].  So the parsimony tally for this hypothesis should be 14,
>not 28, as Keith Robison suggested (Keith seems not to have recognized
>the necessity of event #7, and so missed finding the most 
>parsimonious solution for the out-of-not-Africa hypothesis). 

Whoa! Take a look at the tree for 1-9.  The only common ancestor of all of
these terminal nodes is the ROOT OF THE TREE!  You could salvage your 
argument by spliting 1-9 into 1-6 and 7-9.  However, by doing so you have
won your argument in a way O.Henry would love -- you have constructed an
alternative scenario in which the root of the tree is not in Africa but
in which all paths through the tree at one time pass through Africa. 
Put another way, under such a scenario the common ancestor of all humans
was not African, but all humans have an African ancestor.

0)  Vigilant, L. et al.  1991.  African populations and the evolution of human
    mitochondrial DNA.  Science 253:1503-1507.

1)  McClave, J.T., and P.G. Benson.  1988.  Statistics for business and
    economics.  Dellen Publishing Company.

2)  O'Brien, S.J. et al.  1991. Molecular biology and evolutionary theory:
    The giant panda's closest relatives. In: Warren and Koprowski ed., 
    New Perspectives on Evolution, Wiley-Liss.

Keith Robison
Harvard University
Program in Biochemistry, Molecular, Cellular, and Developmental Biology

robison at nucleus.harvard.edu

More information about the Mol-evol mailing list