 |
 |
 |
 |
FACSLink: We have A FACStar Plus and a FACScan with Consort 32. I recently
hard linked the HP on the FACScan to a Tawainese take-away IBM compatible PC
(286). I installed Kermit on the PC, and FACSLink on the HP. The programs
appeared to have been installed correctly. FACSLink appears on the main menu
of the HP, and I can get Kermit into server mode.
I ask FACSLlink to read files in #11:/4 (my subdirectory containing over
30 list mode LysysII files, which I'm itching to manipulate with Joe's
beautifull WinMDI). The machine whirs, takes a bit of time to look at the
files, but comes up with the message "unable to read files"
Why, oh Why? is my cry.
Wax Sections: I'd like to know if I'm going about setting up the FACScan
correctly. I digest with s. Carlsberg overnight, filter off supernatant (35
um nylon mesh) and discard, add Vindalov's citrate buffer with 5%
DMSO, vortex vigorously, collect the supernatant filtered through another 35
ummesh. I get millions of good looking cells this way. I then dilute to have
about 2 million per ml, and take 200 ul. I hit this with Vindalov's B (10
minutes), then C (PI). Specimen kept on ice in dark until acquisition (I only
learned recently that PI stained samples can be kept overnight, without loss
of CVs - it works, and takes a lot of pressure off my life!). I'm happy up to
here - now for SETUP in LYSYSII.
I set up eight windows
1. SSC vs FSC dot plot
2. FSC histogram
3. SSC histogram
4. FL2 height histogram
5. FL2 Area histogram
6. FL2 Width histogram
7. FL2 Area vs FL2 Width dot plot (doublet discrimination a la BD)
8. ditto density plot
I have a wax section of lymph nodes from the same patient, processed on the
same day, for each tumour section (breast). FOR EACH SAMPLE I adjust my FSC
and SSC until I obtain a reasonable population(s), away from the axes,
looking at windows 1,2, and 3. I then adjust FL2 until the first G0/G1 peak
of the lymph sample sits on channel 200 in the FL2 Area histogram, and around
there on the Width histogram. I change to normal, gate out debris at the
origin (but accumulating some for modelling, when I obtain ModFitLT or
MultiCycle, soon), and collect 10 000 events (generally, my preparations do
nothave much debris). I now run SETUP with the tumour, adjusting FSC and SSC
to obtimize the pictures in windows 1,2, and 3.
Here's my first question. Do I leave the FL2 settings as set for the lymph?
We (I must give my colleague Christo Muller credit for these ideas,
especially because he lurks in this group!) thought that running a lymph node
from the same patient and processed the same way on the same day would be
a neat way of definitively establishing the position of the normal G0/G1
peak for the tumour. But, we see too many aneuploid tumour peaks for peace of
mind, so should we go back to assuming that the first tumour peak is the
normal G0/G1? In which case there's no point of running lymphs.
Next question. If that is the case, should we adjust each sample so that that
first peak falls at channel 200? Our gut feeling is yes, because of the
possilbility of non linearity along the channel axis.
Final question (for now). ISAC guidlines are that a CV of 8 percent is cut
off for wax sections. With LYSYSII one can set the histostats markers to get
whatever CV you desire. We have tried to be consistant by using the
horizontal line between the "goal posts" as our guide. When the intersection
between a goal post and the horizontal line hits the histogram, we click.
Does this give us a "half CV" value? We know that many recommend reporting
one third CVs. Is there a way to do this in LYSYS?
We would appreciate any comments/guidance in these matters. Be gentle, we're
only flow tyros (but we're trying to do it right).
John Michie PhD
Department of Radiation Oncology
Gene Louw Building
Tygerberg Hospital
Cape Town 7505
South Africa
E-Mail: jm5@maties.sun.ac.za
Tel: 027 21 938 4911 bleep 348
-----------------------------------------------------------------------------
|
|
|
|