notageek

“This” being my argument that any selective advantage noncoding DNA regions confer is not because they sop up mutations due to faulty copying mechanisms.

Suppose two genomes of length L1 and L2, in base-pairs. Suppose a mutation rate r, in mutations/bp–say 1/n. Suppose that the composition of L1 is C, a coding region of length L1, and that the composition of L2 is a coding region of length C and a noncoding region of length N.

L1=C
L2=C+N

The mutation rate for each genome is constant at r. The number of likely mutations in L1 is

r*L1 = r*C = C/n

number of likely mutations in L2 is

r*L2 = r*(C + N) = r*C + r*N = C/n + N/n

in either case, the number of mutations likely in the coding region is constant.

The argument made in class was that a genome with no noncoding material is guaranteed to have any mutations happen in a coding region. Ok, duh. That then proceeded to “so a genome with some noncoding material might have the mutation happen in a noncoding region”, and i think that’s bunk. Here’s why:

Suppose you have two genomes, both of length L1, suppose constant rate r = 1/n bp, suppose composition of L1=C1, composition of L2 = C2 + N2.

C1=C2 + N2; C2 = C1 – N2; C2 < C1 for all N2 < 0

mutations in L1 = r*L1 = C1/n
mutations in L2 = r*L2 = r*C2 + r*N2 = C2/n + N2/n

we already showed that C2 < C1, so yes,
r*C2 < r*C1 and C2/n < C1/n, but only because C2 < C1.

Rather; yes, there are fewer mutations in the coding region of the genome that has a noncoding region, but only because that coding region is smaller. The proportion of mutations in the coding vs noncoding regions (in this polymerase-based argument) varies only with the proportions of the lengths of each type.

I am not discounting any advantage extra DNA might confer in different mutational circumstances, like the UV radiation one (one I had already made, I might note), but to argue that a noncoding region will sop up mutations from polymerase errors seems spurious to me.

Addendum:
Someone In class kept saying “look at the ratios”, so let’s look at the ratios.
The proportion of mutations in the coding region will be C/(C + N); as the proportion of noncoding DNA increases, the proportion of (this type of) mutations in the coding region will be smaller, true, but only because as noncoding DNA is added, the total number of mutations increases.