Turbo Sort Recovery

Joe Trotter trotter at scripps.edu
Tue Sep 5 23:29:48 EST 2000


Doing a drop delay profile on a Vantage using Counter mode and then sorting
in Normal-R mode with a 1.3 drop deflection envelope may be asking for
trouble. Counter mode adds 0.5 drops on either side of the drop envelope,
thereby masking the prescision one needs to be able to clock down to a 1.3
drop window. I would suggest using the "3 puddle" method where you sort in
Normal-R using 1.0 drop envelopes at: -1 drop, 0 drop, +1 drop from the best
estimate of the delay running particles at ~ 500/second. Sort 1000 - 1500
events into each puddle and adjust the delay such that you get the same
number of outliers in the -1 and +1 offsets (and the vast majority in the
middle puddle). By using a real 1.0 drop envelope and using the larger sample
size it is far easier to accurately time the instrument to within 0.1 drops.
We have never used traditional drop delay profiles to time the sorters since
the "3 puddle" method is faster and more accurate.

The Vantage SE is a essentially a digital sort system, whereas the Vantage
uses faster peak sense and hold circuits inherited from the FACStar Plus.
Upgrading a Vantage to a Vantage SE makes a lot of sense if you want better
sorting at "TurboSort" performance.


                    Regards,

                        Joe

Marty Bigos wrote:

> BD has put out (at least) two versions of high(er) speed sorting,
> both called TurboSort (at least informally).  The first product was
> an upgrade to what was then the current Vantage sorter put out by
> BD's special products (or venture engineering) group. We ordered this
> when we got our Vantage in 1995 and found the TurboSort upgrade to be
> very touchy to set up and run so we returned this product. BD
> subsequently re-engineered TurboSort and integrated it with the
> Vantage SE. We do not have this product but have heard from other
> facility managers who do (e.g. Simon, below) that it works quite well.
>
> Regarding recovery (Matt's problem) the standard formula using Matt's
> conditions ( see http://facs.scripps.edu/recovery.html) should result
> in 80%+ recovery for a 10% sorted population run at 5000/sec with a 2
> drop deflection envelope, and a 60%+ recovery if run at 10000/sec. If
> the sorted population is a higher percentage of total events, then
> these values will increase, and decrease if a lower percentage.
>
> "Recovery" in this case means % of possible sorted events. The
> Vantage counters should be counting actual sorts, so what one gets
> should agree (more or less) with them. However, whether that is
> really true with the "old" TurboSort option is something I do not
> know.
>
> Marty Bigos
> Stanford Shared FACS Facility
>
> >Hi
> >
> >I run a Vantage SE with turbosort and find that the numbers of cells
> >in the tubes
> >matches quite closely the numbers on the counters. I run everything
> >pretty much as you
> >do. I understand that the counters on the SE do not include aborts
> >whereas the regular
> >Vantage will count suitable but aborted cells as sorted. I had a
> >similar experience as
> >you with our FACStar Plus. Maybe you could get the counters on your
> >instrument upgraded.
> >
> >Simon Monard
> >FACS Lab Manager
> >Trudeau Institute
> >Saranac Lake
> >NY12983
> >
> >Ph 518 891 3080 X352
> >
> >
> > >>> "Matt Wikstrom" <mattew at cyllene.uwa.edu.au> - 8/30/2000 2:47 AM >>>
> >This one's for all you Turbo Sorters out there
> >
> >I have a BD Vantage with Turbo Sort and have found that sort
> >recovery (ie the number
> >of cells that end up in the tube vs the number on the sort counter)
> >can be very poor.
> >
> >I routinely sort at 34psi and 60kHz ddf and the recovery seems to
> >decrease as I
> >increase the sample rate eg 60% recovery at 5000 events/sec and 35%
> >recovery at
> >10000 events/sec.  Of course, the recovery varies somewhat according
> >to the type
> >of cells and the frequency of the target cells, but the trend seems
> >to be the same.
> >The purity is always good though, around 95%, regardless of the
> >recovery.  How does
> >this compare with the experience of others?
> >
> >I always use a drop delay profile to set the drop delay (routinely
> >between 29 and 30
> >drops) when setting up and monitor the side streams and breakoff
> >throughout the sort.
> >I haven't compared the recoveries for the different sort modes or
> >played around with
> >drop envelopes: most sorts are done using 1.3 drops in Normal-R mode.
> >
> >As far as aborts go, I adjust the sample concentration and the dead
> >time to minimise
> >them.  At sample rates below 6000events/sec, the aborts are always
> >less than 1/10;
> >above that, they creep up to 1/9 or 1/8.  I never exceed 10,000
> >events/sec (sample +
> >abort) when sorting.
> >
> >I've fiddled with the phase gating switch on card#6 and changing it
> >from the middle
> >position to the up position doesn't seem to do much to the recovery,
> >although I have
> >yet to test this on the same day, and have left it in the up position.
> >
> >I guess what I'm trying to determine is whether my instrument is
> >behaving according
> >to spec or not.  Any hints, tips, suggestions, voodoo, superstitions
> >you have that
> >might help improve my recovery would be appreciated immensely.
> >
> >Also, I'd like to hear about any experiences you've had with the SE
> >digital upgrade
> >for the Vantage.
> >
> >Thanks
> >
> >Matt Wikstrom
> >Lotteries Flow Cytometry Unit
> >University of Western Australia
> >email: mattew at cyllene.uwa.edu.au




More information about the Cytometry mailing list