image analysis

mwarner mwarner at NOSPAM.dogwood.botany.uga.edu
Wed Feb 24 11:48:02 EST 1999


William Sun wrote:
> 
> Hello all,
> 
> I wish to quantitate my northern signals on autorad film.  What is the best
> and cheapest way?  Can I just scan the film with a conventional flatbed
> scanner and analyze the image with ImageQuant?  Is this accurate enough?
> Thanks for any advice.
> 
> William
> 
> --
> William Sun, Ph.D
> Metabasis Therapeutics, Inc.           Phone: (619) 622-3930
> 9390 Towne Centre Drive                FAX:   (619) 622-5545
> San Diego, CA 92121                    Email: william at alum.mit.edu

William,
	You can do this, BUT you need to take many things into account.  The
previous post about the film response being log. is true, but your
autorads are useable.  The best thing to do is to include a known set of
standards on every autorad.  You scan these then plot the o.d. to
concentration to check and see how (non)linear the response is. Your
unknown sample must fall within this standard curve as well.  This is an
internal calibration, and you can also do an external one with a
transparent step tablet.  This brings me to my second big point, you
should always scan your autorads in transparency mode, not reflective!
Finally, you can also pre-flash your film before you expose it to your
blot. This is was done quite a bit when people would use x-ray film for
fluorography. You can find a method for this in books like "Gel
electrophoresis of proteins, a practical approach" eds. BD Hames, D
Rickwood, IRL press, 1981.  Lastly, it also helps to use an
auto-developer, you probably are using one, but it doesn't hurt to throw
that in too.  Good Luck

Mark Warner

(remove the spam before e-mailing)




More information about the Bio-soft mailing list