DNA error rates

Robert Bradbury rbradbur at hardy.u.washington.edu
Mon May 18 14:04:00 EST 1992

In article <robison1.706032179 at husc10> robison1 at husc10.harvard.edu (Keith Robison) writes:
>Go back to stat class.  If the probability of losing one copy of SOD
>is 2*10^-5, then the probability of losing both copies is that value
>_SQUARED_, not halved!  So your final fraction is 4*10^-10 (a mighty
>small number).  I don't have the figures in front of me, but 1 in
>40 trillion would be much less than 1 cell/person.

Keith is right.  Probability is not my strong point, my misadventures
at roulette tables due to my inability to comprehend the laws of probability
are ample testimony to this fact.  :-)  At least someone understands the
argument enough to correct my errors.

However, 4*10^-10 as an two allele SOD knockout rate is about 1 in
2.5 *billion*.  I'm fairly sure we have many times that many cells.
(Does anyone have a good ref. for the #'s of cells in various organs
in the body??)

However this does bring to light the glaring difference between the
effects of mutations in the X or Y chromosome vs. mutations in the autosomes
if we presume that everyone has 2 good copies of all their autosomal
genes.  Now I suspect that this assumption is very iffy for several
reasons.  there are a large number of known genetic diseases
(4000+ and counting).  Some significant portion of these must be
recessive diseases meaning we can have lots of carriers in the population.
The carrier level of CF is one in 25 whites.  I've seen figures for
others in the 1 in 100's to 1 in 1000's.  (Anyone have a list of carrier
frequencies?)  If you add all of these up there may be quite a few of
our autosomal genes for which we do not have 2 "working" copies.

Now, looking at the "Parameters of the Human Genome", NE Morton, PNAS 88:7474-6
we can see the X chromosome is 5.1% of the genome.  Assuming that we have
100,000 genes and that one in 10 are "active" in a cell, we must have 500
active genes on the X chromosome.  Going back to my 2.2*10^-5 mutation rate
we get 1 in 100 cells with a broken X-chromosome gene.  Not my initial
one broken gene / cell claim but this doesn't include any mutations
in autosomal genes for which you don't have two good copies.

Aging is closely associated with the loss of cells in many tissue
types.  While at 80, the loss may only be 20-30%  (except as has
been pointed out in Parkinsons where it is 80% in the SN) it is
a downhill slide with the numbers reaching into the 50%+ range
in many tissues around 150.  Now given the observed cell death
increase and given the observed mutation rates and the mutation
rate increase with age, I want someone who doesn't believe that
DNA damage is a big factor in aging to explain why.

More information about the Ageing mailing list