Re: Percentage values commonly given for differences in DNA etc, may be meaningless.

Jane Andrews (
Mon, 29 Jul 1996 16:37:51 +0100

On Fri, 26 Jul 1996, Gautam Majumdar wrote:

> In article <>, =?iso-8859-
> 1?q?Ludvig_M=F6rtberg?= <> writes
> >Is there really an established way of measuring differences between
> >two DNA sequences that gives you a percentage value? If we take for
> >example a 100 bases and substitute 2, this may be a 2% difference. But
> >what if we insert 5 bases or delete 10, then what?
> Also how can we tell the total percentage difference of DNA between
> two species without knowing the complete sequence ? Complete
> DNA sequencing, as I understand, has been done only for some
> bacteria and yeast.
There is a way of comparing the entire DNA compilment of 2 species which
is along the lines described earlier in this strand known as DNA-DNA
hybridisation. The genetic compliment prepared from 2 species are heated
to allow the double stranded structure to seperated. Then they are mixed
and slowly cooled; the DNA will reanneal. In some cases the DNA will
joint up with the opposing strand from the same species and in some cases
from the other species. The DNA is then heated again and the temperature
at which it denatures is noted. A comparison of these temperatures
recorded from the pairwise comparison of various species can be used to
calibrate the genetic similariy of any pair. The more similar that the
DNA of two species is then the stonger the bonding between the parallel
DNA strands and so the more heat that will be needed to seperate them.

This is the basic principle, anyway. Details should be available from any
molecular genetics text book. The hybridisation process compares DNA on a
gross level and I don't imagine that it's sensitive enough to be affected
by a 5bp insertion, but given the number of base pairs in the entire
genome of a species then 5 or 10 are very few.

Jane Andrews.