[LMB] No lefties

Thomas Shaw topaz0 at gmail.com
Sun Oct 6 03:31:32 BST 2019


On Sat, Oct 5, 2019 at 8:17 PM Joel Polowin <jpolowin at hotmail.com> wrote:

Perhaps I should have said that the error rate is "high" rather than
"bad".  Though most such errors *are* harmful, at least for the
individual.  A random change to a really complicated finicky system is more
likely to leave it in worse shape than before, instead of better.

Thomas: This isn't my field exactly but it's tangential to it, so I do hear
talks on the subject from time to time. I get the impression that
geneticists/evolutionary biologists argue a fair amount about the relative
frequencies of deleterious, neutral, and beneficial mutations. The neutral
ones are interesting to me: a lot of (DNA) mutations are synonymous, i.e.
code for the same protein. A lot of others seem to not do much.

Also, it's common practice in biology research to generate several or many
mutants of a protein of interest to see what will change (not to mention
producing viable mouse lineages that bear these mutations). Sometimes in a
targeted way but sometimes in a brute-force way: sequentially replace each
amino acid with alanine, or pairs of amino acids with cysteines, which can
then be cross-linked for other purposes, etc etc. Generally a lot of the
mutants are viable and functional, and can even be expressed in viable
mice. This surprised me when I first learned about it, but there are also
various reasons it makes sense. Complexity is one of them: biological
systems are filled with redundancies and feedbacks, which is probably an
evolved property -- organisms need to be able to produce the same results
from a wide variety of unexpected external conditions, so robustness is a
necessary characteristic.


More information about the Lois-Bujold mailing list