Phosphoroimagers vs Instantimager Experiences Saught

Peter Heifetz phef at
Tue Oct 4 16:02:11 EST 1994

In article <1994Oct4.155732.19538 at>, brianf at (Brian
Foley) wrote:

> 3)      We have been happy with the software for both the Betagen and the
> phosphorimager.  The phosphorimage comes in either Mac or Windows flavors.
> The phosphorimager software was updated recently and is quite improved, 
> so if you saw a demo a few months ago and didn't like it, you might try 
> again now.

Molecular Dynamics' new software for the NT workstation ("ImageQuaNT") is
somewhat a case of "3 steps forward and two steps back".  Overall it _is_
an improvement; however, some serious deficiencies were introduced
relative to the previous ImageQuant 3.3 program.  Specifically, the "sum
above background" quantification option for volume integration was
removed, as was support for the analysis macro program MultiQuant (a
tremendous timesaver for those of us doing repetitive gels with many lanes
and bands).  Any quantification of images with significant background in
the lane (e.g. Northerns or in-vivo labeled proteins on SDS-PAGE) requires
"sum above bgnd" because of the very low signal between lanes.  If
selection objects span this area (as they must if the object is to enclose
the band of interest without contacting it or a neighboring band) and the
within-lane background signal is correctly used as the "background" for
integration, the between-lane pixel density will be a negative number when
straight "volume integration" is used, and will be subtracted from the
integrated pixel density of the band.  You can actually end up with
negative "counts" for a prominent, dark band (!).  Sum-above-background in
IQ 3.3 truncates all pixels below the set background to zero, the only way
to correctly handle the "two-level background" situation in these gels. 
In the new ImageQuaNT there is a quantification option deceptively called
"Sum"; this treats the _background_ as zero and merely sums the total
pixel density with no background subtraction at all.  I've had lengthy
phone conversations with Molecular Dynamics about this.  According to
them, the omission was inadvertant and they will be returning "sum above"
to the _next_ version of ImageQuaNT.


>         We use a tape drive (6 gigabytes per tape, I think each tape 
> costs about $15) to back up and store data.  The phophoimager hard disk 
> was partitioned to have software on one partition and user space on 

If you can afford one (ca. $700) a magneto-optical drive is great.  128 Mb
disks cost about $30 each, and in contrast to tape are fully random-access
and as fast as a slow hard disk (and much much faster than floppies).  You
will need a SCSI card for the PC, but they're not too expensive.

Peter Heifetz
Duke University
Developmental, Cell, and Molecular Biology Group

More information about the Methods mailing list