Re: [MLUG - DISCUSSION] these are the days of miracles and wonders: a radically new sequencing machine appears

[Stephen Montgomery-Smith <EMAIL:PROTECTED>]

> I'm glad that you've
> appreciated that this is subtle stuff.

The real reason I am getting interested in this is because recently I 
have started talking with some biologists about statistical problems in 
these areas.  I have recently started attending a seminar held on 
Tuesday mornings in Tucker Hall, that is led by George Smith.

He developed a technique of trying to recognise "receptors" in 
cells/antibodies/whatever by putting short sequences of protein 
(ligands/peptides) onto the ends of thin little viruses (called phage 3) 
and exposing them to the substance with the receptors, and seeing which 
ones stick.  (They send the ones that stick to a lab where they get 
sequenced.)

The way he creates these short sequences of peptides on virus's is by 
splicing short sequences of randomly created DNA into the DNA sequence 
of this phage, and then putting this DNA into E. coli (the bacteria(?) 
commonly found in our stomachs), where they do what a virus does, and 
replicate themselves from the E. coli.

The seminars are fascinating.  A recent one was where they use these 
peptides to detect tumors.  There is some peptide they have discovered 
using this technique that attaches to certain types of tumor.  They put 
these peptides into a "designer protein" that also includes a 
radioactive isotope.  They squirt this into mice, and then look for 
where the radioactivity comes from.  But it is all rather tricky, 
because most of this radioactive stuff ends up in the kidneys and that 
kind of place.  But there is a faint glow in other parts of the body 
where the tumor is.  (They have to have these mice on the radioactivity 
detection machines for several hours because the signal is so weak.)

Another seminar was about where they had looked at peptides that stuck 
to antibodies of people who had limes disease.  Limes disease is 
apparantly a very difficult disease to diagnose, in part because there 
is another tick carried disease in Missouri that has exactly the same 
symptoms.  Anyway, this professor looked at the peptides that had stuck, 
and looked at the common "motifs" in a database of well known proteins 
to see what he could find.  Interestingly enough, one of the motifs 
looked like it was for strep-throat.  Apparently what he had found was a 
correlation between people with limes disease (or rather, people 
diagnosed with limes disease) and people with strep-throat.  But he did 
find some other motifs that did look like real limes disease (these are 
called epitopes, I think).

After this seminar, it made me wonder just how doctors can ever reliably 
diagnose any disease.  Does anyone know how the rapid-strep test works? 
 Is it very reliable (like 99.9%) or just moderately reliable (like 85%)?


_______________________________________________
discussion mailing list
EMAIL:PROTECTED
http://mlug.missouri.edu/mailman/listinfo/discussion