DiVa - counts and cells numbers

Hi all,

We have a series of question about event rates and cell and sort 
counts on the FACSVantage DiVa. Over the last few months we have 
observed the following:-

* EVENT RATES - the rate shown on the analog CRO and the in DiVa 
software are sometimes similar but can often vary by up to 20% (with 
the analog CRO nearly always being higher). We suspect that some of 
this difference is due to different thresholds in the two "sides" of 
the machine but I'm not convinced this is the whole story as varying 
the threshold in the DiVa software does not always have any influence 
on the difference between the rates.

It has been suggested to us that this variation is normal. However, 
reading through the DiVa 4.0 manual recently I came across the 
troubleshooting section (p229) which suggests a low event rate (ie 
lower than "threshold counter on the instrument" (which I assume 
applies to the analog CRO?)) can be due to "memory full".  We tried 
the suggestion of restarting a couple of times but it didn't seem to 
make much difference. The implication is that the two counters should 
match...

QUESTIONS - do other people see this? Is it normal? Has anyone 
identified factors other than threshold that can contribute to this 
observation? And, most importantly, does it matter? ie which is the 
important number to watch when considering how fast to run a sort and 
could this be contributing to the issues described below...


* NUMBER OF THRESHOLD EVENTS - some of our scientists have started 
doing very careful counts of the number of cells they are bringing 
down to sort. They are then comparing those numbers with the total 
number of processed events in the sort report. In some cases the DiVa 
is processing only ~50% of the number of cells the scientists believe 
they had in the tube before the sort.

Now, obviously there are lots of variables that could be at play 
here, but, for example, last week we did two sorts on the same 
morning for the same scientists with two samples that were prepared 
at the same time. The first sort tube supposedly had 48x10^6 cells in 
it and the DiVa processed ~23x10^6 events during the sort. The second 
tube (which contained the same cells but with a different stain), 
reportedly had 18x10^6 cells and DiVa reported processing ~16.7x10^6 
events.  This sort of variability is what has us concerned.

QUESTIONS - again, is it 'normal' for the DiVa to report processing 
much a lower number of events than what are reportedly in the tube 
(ie 50% lower!)? If so, why does it sometimes seem to get it right? 
Any ideas on what might be causing this?



* SORT COUNTS and CELL RECOVERY - we are also seeing huge variability 
between the sort counts and the actual number of cells that are 
recovered. Again, on the same morning mentioned above the first sort 
claimed to have ~2x10^6 cells in the left tube and ~4x10^6 cells in 
the right. The researcher recovered only about ~50% of this number. 
However, in the second sort the machine reported ~1x10^6 cells to the 
left (only a one-way sort) and this tallied very closely with the 
post-sort count. (I wonder if there is anything in the fact that the 
first sort is so far out in all respects whereas the second was much 
closer?). Both sorts were done with the standard "Purity" mask in the 
DiVa 4.0 software.

QUESTIONS - again, is this normal? We are aware that reported numbers 
of sorted events are often higher than true counts (certainly this is 
the case on our FACStarPlus), but the variability is again what 
really concerns us. Do others see the same sort of variability? Does 
anyone have a handle on what factors can lead to this sort of 
variation?

When trying to sort for target numbers (eg when doing a multi-stage 
sort where we are trying to obtain multiple populations from the same 
sample) then this variability makes it very difficult to know when to 
stop and proceed to the next population...


Any suggestions and comments would be most welcome!

Regards,

Adrian


-- 
______________________________________________
				Dr Adrian Smith 
		    Research Facilities and IT Manager	     
       Centenary Institute of Cancer Medicine & Cell Biology	   
	   Locked Bag No.6 Newtown, NSW 2042 
AUSTRALIA.									
	 Ph: (02) 9565-6189	Fax: (02)-9565-6101