3700 contamination among capillaries

Sven Klages klages at molgen.mpg.de
Mon Aug 6 02:10:21 EST 2001

Hi Babara,

there are others with the same problem .. :-(
With shotgun sequencing projects you probably won't
get any problems (due to the philosophy of such
sequencing approaches).

When using 3700s in a sequencing core or when running
inidividual samples (finishing reads for shotgun projects
or cDNAs) this behaviour is pretty dangerous.

We found even with signals as low as 300-400 "bleeding
capillaries". In every case spatial calibration is okay,
and even with new probetips the problem did not disappear.
We were told to raise the wash volume in the run modules
from 200 -> 600. This may be a good idea but it didn't
really help.

We observed, that the 'wrong' sequence has an extremly
low signal (though the traces sometimes look really good);
so it might be a workaround assuming that traces below
a certain threshold (e.g. signals below 50) are artefacts
and to delete them.
Comparing each capillary with its neighbour with e.g
'phred'/'cross_match' and then eliminating suspect data 
works quite fine in our hands.

So this probably won't work in every case ...
There is no solution and no statement from Foster City.
It seems that this problem is non-existent for ABI.

On the other hand there should be enough "space" in the
spatial dimension, that this phenomena should not appear
(in theory ;-)

Any other folks with the same problem .. ?
Are there any solutions ?

Maybe there are some people who never checked or observed
if there is that kind of problem .. !?

best regards


On 4 Aug 2001 14:16:12 +0100, bmr <bmr at cribi.unipd.it> wrote:

>I have a contamination problem among capillaries in the 3700 and I was
>wondering if there is someone else that has the same problem and maybe
>has solved it.
>I discover it because in a well where there should have been no segenal,
>I could read a sequence.
>Later I discovered that the problem was also in other capillaries.
>The contamination is among adjacent capillaries
>Es: cap. 71 had the same sequence of cap. 69
>The sequence in cap 71 was 300 bases and sequence in 69 was double for
>the first 300 bases after while it continued as single.
>In the double portion I could clearly recognize the sequence in cap.71.
>I could also see the signal from cap.71 in cap 70.
>I have many other case like this.
>I have been told that the problem is due to the fact that the signal of
>contaminating sample is too high and in effect many of them where out of
>range (>7000), but I made another run of the same 96 plate with a lower
>injection voltage and the signal of contaminating samples resulted
>1000-3500, but the contamination was still presents even if in less
>capillaries than the first run.
>I made a new spatial calibration before the second run.
>Could someone help me?
>Thsnk in advance
>                Barbara Simionati
>CRIBI Biotechnology Centre
>  and Department of Biology
>  University of Padua, Italy
>  Tel. 049-8276160


More information about the Autoseq mailing list

Send comments to us at biosci-help [At] net.bio.net