1998-10-14 - RE: DNA

Header Data

From: “Brown, R Ken” <brownrk1@texaco.com>
To: mmotyka@lsil.com>
Message Hash: f471573e07a5561be316210f2fc802894787f93688cc224ec68bae33c0cda23a
Message ID: <896C7C3540C3D111AB9F00805FA78CE2013F84D3@MSX11002>
Reply To: N/A
UTC Datetime: 1998-10-14 12:41:46 UTC
Raw Date: Wed, 14 Oct 1998 20:41:46 +0800

Raw message

From: "Brown, R Ken" <brownrk1@texaco.com>
Date: Wed, 14 Oct 1998 20:41:46 +0800
To: mmotyka@lsil.com>
Subject: RE: DNA
Message-ID: <896C7C3540C3D111AB9F00805FA78CE2013F84D3@MSX11002>
MIME-Version: 1.0
Content-Type: text/plain



Mike Motyka wrote: 

> I'm not a biochemist 

Neither am I but I'm going to classes... 

> I think modern DNA analysis yields data that bear more 
> resemblance to a spectrum than anything else. In this 
> case there is always a good possibility for error.
>  Remember how large a project the Human Genome
> thing is? That is an actual sequencing project.

AFAIK police & courts don't  use actual sequencing but DNA "typing" or
"fingerprinting" - they extract certain well-known genes, a tiny subset
of all that are there, and compare them with reference samples. You
don't have to actually  "read" a gene to be able to match it with
another one.  The genes they used aren't ones that are expressed either.
They are if you like the "comments in the code".  Or more likely
graffitti because they don't seem to mean much.

You can't "prove" that a sample did come from a named individual, but
you can prove that it *didn't*. Which is why DNA testing has been
popular with defence lawyers in this country (UK) because it has been
used by the defence to get people released in a number of famous trials.
(There was one poor bloke who got let out of prison after over 20 years
when DNA typing showed he couldn't have committed the crime for which he
had been convivted. He hadn't been released on parole because to get
parole you have to show some contrition and he always claimed he was
innocent).

You can give a likelihood that a sample came from the same person as
another sample though. Or their twin, or parents et.c et.c. It's also
been used to get past our notoriously racist immigration officials. They
tend to assume that anyone coming into this country as the spouse or
parent of a citizen is lying (especially for Asians - they assume that
all marriage are arranged or "marriages of convenience" - for some
reason white Americans and especially Australians don't have nearly so
much trouble.). DNA typing has been used by immigrants to show that they
are (probably) the parent.

http://www.dartmouth.edu/~cbbc/courses/bio4/bio4-1997/KatieLachter.html
has some bumf about DNA evidence in US courts (you are much more
suspicios of it than we are)

http://www.emf.net/~iisme/actionplans/201.html has instructions for
do-it-yourself DNA typing! 
Unfortunatly  it requires equipment you are unlikely to have at home (&
do US universities  *really* issue such detailed instructions for lab
work? "Raise your hand. Move it over the pencil. Move it down until the
fingers are in contact with the pencil.  Close your fingers round the
pencil. Why is it important not to exert too much force with the
fingers? Pick up the pencil. Remember the point is sharp and may have
been exposed to potentially carcinogenic solvents. Do not stab any part
of yourself or any other student. Do not place the point near your eyes,
mouth or genitalia. Do not use the pencil in the presence of any minor
under the age of 18, or any vertebrate experimental animals, pets or
livestock. Do not hand the pencil to anyone not covered by the college's
insurance scheme or who has not signed the standard disclaimer form..."
OK, I exagerrate. But not much.)

Within 20 years techniques that are now restricted to university labs
and pharmeceutical companies will be available to the hobbyist. Maybe
bio-hacking will be as influential in our kids lives as computer-hacking
has been in ours.

> And pattern analysis will identify the guilty before they 
> even think about committing an offense. 
> Hell, they'll probably be able to ID the bad ones from 
> samples coerced/stolen at birth and then all future
>  opportunity and resource allocation will be effectively 
> decided by the government, inc.

Again, current methods don't identify genes just show how much they
resemble those of another sample. And they mostly use DNA that is not
transcribed anyway. So they tell you nothing about what a person looked
like.  Of course the stuff coming out of the genome project (which is
having the same effect on molecular biology that the cold war had on
rocket science, or the discovery of America on navigation, or the
railways had on steam technology) will probably change all that Real
Soon Now.  There are at least 3 or 4 new methiods that are being
introduced this year.

> Could it be that a single sample of DNA can be used
> to produce a large supply of that DNA which could then 
> be used to fabricate evidence? 

PCR can multiply the size of a sample many thousandfold in a day.

Ken Brown





Thread