ortholog and paralog

Laurent Duret duret at biomserv.univ-lyon1.fr
Mon Feb 17 03:42:18 EST 1997

In article <3303e13d.0 at>, pryor at mail.teclink.net writes:
>duret at acnuc.univ-lyon1.fr (Laurent Duret) wrote:
>>   "Where the homology is the result of gene duplication so that
>>   both copies have descended side by side during the history of 
>>   an organism, (for example, alpha and beta hemoglobin) the genes
>>   should be called paralogous (para = in parallel). Where the 
>>   homology is the result of speciation so that the history of the
>>   gene reflects the history of the species (for example alpha 
>>   hemoglobin in man and mouse) the genes should be called orthologous
>>   (ortho = exact)."
>OK, fine. But how do you know the beta globins are homologs? Could you not 
>have duplication of beta globins in an ancestral species, with one lineage 
>maintaining functionality in one and the other the other..with genomic 
>rearrangements, etc. how could you tell..or due you assume that there was 
>no duplication of beta globins.. an assumption which appears to be 
>absolutely incorrect.

You're right! It is possible to prove that two genes are paralogous,
but it is always difficult to demonstrate that two genes are orthologous...
One may say that two genes are orthologous ... as far as one has not
proved that they are paralogous.

Take the following example: one has sequenced 3 homologous genes, in
man, mouse and xenopus. After alignment, we obtain the following

         ____________ man
  |     |____________ mouse
  |___________________ xenopus

You may assume that the man and mouse genes are orthologous.
Later, another homolog from chicken is sequenced, and it gives
the following tree:

                _______ man
   _____|      |______  chicken
  |     |____________  mouse
  |___________________ xenopus

In this tree, man is more closely related to chicken than it
is to mouse. This obviously contradicts what we know about vertebrates
phylogeny. Hence, we can conclude that a gene duplication occured
before the mammals/birds divergence. Thus in this case man and mouse genes 
are paralogous, whereas we first had thought they were orthologous.

Note that the relative length of the branches that separate
man and mouse and xenopus in the first tree might have alerted
us of the risk of paralogy. However, since there is no absolute
molecular clock, the branch-length criteria cannot always help
in determining ortho/paralogy relationships.

Indeed, the more genes we have, the more confident we can be in
our prediction of orthology.

>How could you ever possibly know that "a gene reflects the history of a 
>species" without making a prior assumption that the gene you want to 
>"reflect the history of the species" does reflect the history of the 

I'm not sure to understand your remark.
In many cases, we have information about species phylogeny,
independantly of genes (fossils, comparative anatomy, etc...).
I such cases (as above) we can use the 2 independant information.

In the case where we have no idea about species phylogeny, of 
course it will be difficult to determine ortho/paralogy. 
However, as more genes are sequenced, it is possible to compare
several gene trees to infer the most likely species tree.

Best regards,

Laurent Duret

Laurent Duret                           
Laboratoire BGBP - UMR CNRS 5558     Phone  : +33 472 44 80 00  p.34 39
Universite Claude Bernard - Lyon 1   FAX    : +33 478 89 27 19
43 Bd du 11 Novembre 1918            e-mail : duret at biomserv.univ-lyon1.fr
F-69622 Villeurbanne Cedex           ========================

More information about the Mol-evol mailing list