Hi,
I agree with David - I've had an array crash out at ~100 runs. Data was
good at first, then started to get very patchy (individual capillaries
dropping out) then finally the data was useless from the entire array, even
with the pre-sequenced pGEMs that ABI send out on install. It took about 10
runs to go from good data to 0 bases from the entire array. Once the array
was changed (keeping same lot of POP6, buffers etc) we went up to 700bp from
our own standards (pGEM again).
We do a nitric acid wash every 100 or so runs as we do notice a drop in read
length at this time...but it can be as soon as 80 runs or as long as 130
between washes. We tend to go off our own pGEM sequencing when timing the
wash as I know exactly how much data we should be generating form these...as
usual our users data tends to vary more in quality than our standards.
BTW - ABI recommends binning the POP6 after 5 days on the instrument...has
anyone tested it beyond that...I get 7 days with good data.
Stuart.
Stuart Bayliss,
Manager,
Genetics Core Facility,
MRC Clinical Science Centre,
Hammersmith Hospital,
Du Cane Road,
London.
W12 0NN
Tel (lab): 020 8383 3181
Tel (office): 020 8383 8305
Fax: 020 8383 8338
email: stuart.bayliss@csc.mrc.ac.uk
web: http://gcf.csc.mrc.ac.uk
-----Original Message-----
From: David Mohr [mailto:dwmohr@mail.jhmi.edu]
Sent: Thursday, March 15, 2001 2:00 PM
To: Recipients of ABRF List
Cc: tjs11@cornell.edu
Subject: Re: big dye lot?
Hi All,
These problems sound all too familiar. In our experience, 9 times out of 10
the problem can be traced to the array or an instrument related failure
requiring a service call. Some of our arrays have only lasted ~100 runs.
Part of this is probably due to the junk that we inject in our capillaries.
In our lab, a core facility that runs other peoples' samples with a varying
range of quality, we see a definite degradation of sequence read length over
time, even though the ABI standards work quite well. Sometimes regenerating
the array helps, but more often than not, we have to replace the array. I
think part of the blame lies with the samples themselves, but much of it has
to do with the variability of array quality. We've had plenty of problems
just getting an array that works. I think a testament to the whole picture
is how readily ABI will replace arrays without charge.
One of the ways we test our array's quality is by setting up pGEMs ourselves
and comparing them over time. In our hands, this will show an array problem
much earlier than the ABI standard.
Let's hope it's not the polymer lot...we have been using POP-6 lot 0101130
for the past 3 weeks and haven't noticed a problem.
David Mohr
DNA Analysis Facility
Johns Hopkins University
Baltimore, MD 21287
----------
>From: Thomas J Stelick <tjs11@cornell.edu>
>To: Recipients of ABRF List <abrf@aecom.yu.edu>
>Subject: Re: big dye lot?
>Date: Wed, Mar 14, 2001, 12:40 PM
>
> bob,
> the problem we have seems similiar. We had some real bad
problems
> with a new array we installed in october. After beating our heads against
> the wall for a few months, we were able to get a confession that some of
> the arrays from abi had some problems. After we replaced the array in
> january, things started going better. But over time we have been seeing a
> slow degradation. A lot more noisy sequences and more outright
> failures. We have run the abi bid dye standard and the machine appears to
> be working correctly. (although if you have used these before the signal
> strengths are about 20 times the signal you get from a normal sequencing
> reaction). It has recently gotten much worse. We have either a bad lot
of
> polymer, bad big dyes or a different chemical problem. It still might be
> an array issue though. We just did a regeneration and that did not seem
to
> correct the problem. We are at about run 100 on this array. I have
> ordered in a different lot of big dyes to try that. Hopefully this helps
> and I can stop beating the wall.
>
>
>
>
>
> At 09:06 AM 3/14/01, you wrote:
>>Thomas J Stelick wrote:
>> >
>> > hello,
>> > We have been getting some inconsistant results on some of our
sequencing
>> > and one problem may be the lot of big dye. The problem showed up
around
>> > the time we started using a new lot. The lot # is 0101047. Is anyone
else
>> > having problems. We are using a 3700. Or is anyone having a problem
with
>> > pop-6 polymer lot # 0101130.
>>
>>
>>We have been questioning POP6 lot 0101130 ourselves. An instrument that
was
>>working very well with read lengths of 700-800 suddenly changed to reads
of
>>350-450 after only a POP6 lot switch. ABI had no other reports of a
problem
>>with that lot.
>>
>>There were reasons to suspect other problems: perhaps a bad array
(regenerate
>>array improved the situation and switching arrays got us up to 650 nt
reads)
>>and perhaps an instrument problem (we started seeing strange drop-outs in
the
>>array image data in caps 90-96).
>>
>>The symptoms after the POP lot change were (i) decreased resolution -
broad
>>run-together peaks, (ii) *leading*-edges of peaks had reverse "tails" (did
>>that make any sense? I can't figure out how else to describe it) and (iii)
>>numerous capillaries suddenly gave no resolution at all. With such a
routine
>>procedure as POP lot change, I can't image what we could have done that
would
>>have toasted the array. I am still highly suspicious of the POP6 lot.
>>
>>To re-iterate, yes we were suspicious of POP6 Lot 0101130, but no, we
could
>>not prove it was a bad lot, and it is currently giving us modest
resolution
>>but NOT great. I would be very interested in hearing of others using that
>>lot of POP6 and what results they are achieving.
>>
>>Bob Lyons
>>University of Michigan
>
> Tom Stelick
> DNA Sequencing Facility manager
> Bioresource Center
> Cornell University
> Ithaca, NY 14853
> (607)254-4857
> http://brcweb.bio.cornell.edu
This archive was generated by hypermail 2b29 : Mon Apr 02 2001 - 10:20:49 EDT