[LMB] Gene cleaning in real life

Karen Hunt huntkc at gmail.com
Sat Nov 11 13:41:52 GMT 2017

On Fri, Nov 10, 2017 at 7:05 PM, A. Marina Fournier <saffronrose at me.com>

> On Nov 09, 2017, at 06:56 AM, Sylvia McIvers <sylviamcivers at gmail.com>
> wrote:
> How you you get rid of recessive genetic diseases? Tay-Sachs disease
> killed about 1/3,600 babies a year among Jews of European descent... can
> scientists cure it?
> They can't (yet), but the community can advise all high-school seniors in
> religious schools to get gene tested. The lab results help make sure 2
> carriers don't get married. No more dead babies.
> Marina:
> I gather it isn't tested for at birth. If the family contains "hidden
> Jews" in its background, they would not think to get tested, and the
> advisor for genetic testing at amniocentesis or chorionic villius sampling
> (which is usually done a couple of weeks earlier) would not necessarily
> suspect it, nor test for it. If a pregnant woman is under 30, such testing
> is rare.
> I think yours is a good idea, but I think universal gene testing at the
> time of a teenage vaccine or booster would make sense.
> I have one mutation of which I'm aware, but wouldn't have been
> automatically tested for..

A true universal gene testing for "all" conditions of interest is going to
be a while in coming about:

I was genetically tested a few years ago for late-diagnosis Cystic Fibrosis
(no I don't have it, my lung troubles have different causes). The test,
just to examine that one gene, was very complex and expensive. The reason
is that CF can be caused by any of thousands of mutations to that gene, and
in fact the disease takes a variety of severity of troubles and other
details depending on what sort of result that mutation produced. To test
for it, they have to look at each allele of the gene, which is a pretty
large one, and match it to a known CF-causing sequence. If someone has a
variation that isn't in the database, they can't determine whether it's a
natural variant or a problem one. Someday protein folding programs might
get good enough to answer that, but for now it's a huge effort to figure
out one large protein in enough detail to answer questions about the
particular sequence of amino acids, and there's only recently beginning to
be an ability to say that if this sequence is almost the same as that one,
then this will be the change that results.

I don't know how many conditions have thousands of variants like CF does,
but I bet some of them do. For what it's worth, being heterozygous with CF
and not CF appears to give immunity to Yellow Fever. Any reasonably common
recessive condition will have some benefit to being heterozygous, that's
what allows it to become reasonably common.

Karen Hunt

More information about the Lois-Bujold mailing list