I would have thought that DNA sequence chimerism in humans is mainly limited to blood cells, as is commonly thought. (Or assumed by those who don’t work on it first-hand!) Incidentally, I mentioned this in my old post “Monkey business or is my Uncle also my Dad’: http://sciblogs.co.nz/code-for-life/2009/10/13/monkey-business-or-is-my-uncle-also-my-dad/

Do you have a reference for your claim “the evidence is now clear that mosaicism is common and possibly the norm”. (It’s hard to keep abreast of everything. From a very quick look in PubMed, I can see “incidental” papers on this, but not a “key” paper. I can vaguely recall something in Nature, but that’s not much use!)

I’m not sure this is an issue in the way I think you’re thinking though. Long story short, it seems to come down to simply using the same tissue in the diagnostic test as the test was developed on, or if using a test developed on one tissue and but used in diagnosis using another, bearing this in mind.

As long as the same tissue is used as the test was developed on, it’s not an issue. The issue seems to me to be transferring test developed on one tissue to diagnostics based on another, not the mosaicism itself.

It will mean that some (rare) disorders that are dependent on mosaicism in specific tissues and only detected this way might be missed, but this should simply mean that there is in effect no test for this disorder, rather than “confused” tests.

Similarly, if the test are unable to detect some types of (somatic) variation then they’d lack predictive power rather than be “confused”.

Which is to say, it could be a limitation in developing tests, but I’m not sure that it’d affect tests once they’re developed, provided the groups doing the diagnostic screens take care to note this when using the diagnostic on a test based on a different tissue that the initial screen was developed from.

I hope I’m being clear enough.

There is the “opposite” problem of mosaicism the sampled tissue (associated unrelated disorders to that being tested) in “throwing off” the tests , which I’d like to think amounts to “just another” source of noise in developing the test. (If it’s that common that it’s a nuisance it ought to show up, etc.)

Or are you thinking of something else again?! :-)

While I’m writing, another thing I’d watch out for is what’s measured in the research reports: differences occurring at the RNA level may tie in with epigenetic heterogeneity rather than DNA sequence mosaicism.

http://scienceblogs.com/pharyngula/