phred training and capillary questions

C. Desmarais cid at
Wed Oct 27 04:34:09 EST 1999

On 26 Oct 1999, Stephen R. Lasky wrote:

> If phred really needs more training, maybe those of us who use 3700's
> need to start putting pressure on PEBiosystems to supply Phil Green and
> his team with an instrument to train phred on (what chance is there of
> that?). I know he already has one of the other flavors of capillary
> machines.

We actually just received a 3700 for a 6-month trial period, to provide
data to Phil & his group for analysis. In fact, the shrink wrap just came
off about 3 hours ago and the electricians are installing the proper
wiring at the moment. We are hoping to start it running right away, so I
should have some answers regarding phred's basecalling accuracy very

> In this same vein, I was wondering what kind of experiences people with
> capillaries are having? We have two 3700's here and have had abysmal
> uptime and pass rate records on them (let's be generous and say 75%
> uptime (and that's being really generous)).  Has anyone had the
> opportunity to run 3700's alongside MegaBACEs?  How do you feel about
> the MegaBACE?  What kind of life span are people getting out of the
> capillaries from both machines? Have other people done an estimate on
> the cost per lane or cost per Q>20 base from the capillaries as opposed
> to 377s? If there are problems with capillary clogging on the MegaBACE,
> do you think its is because of your DNA prep, and how have you modified
> it to accommodate the capillary machines.

We have a megabace on a 6-month trial period as well (for the same reason,
to provide test data to Phil Green) and it has been a super machine for
us. We have been running it a few times per day with <10% failure rate and
quality averages of 400-500 Q20 bases per sample. We are playing around
with the parameters a bit to try to get higher quality data from it. When
I get a proper report on the two machines side by side, I can pass it on
to anyone who is interested. 

Cindy Desmarais
Finishing Coordinator, University of Washington Genome Center.

More information about the Autoseq mailing list