Automated DNA sequencing - rebuttal

Rick Wilson rick at GENEMAN.WUSTL.EDU
Wed Jun 19 13:32:13 EST 1991

I wasn't going to say anything but after reading Donald Lehn's comments, I
feel I must ...

First of all, the "370A" isn't "...the current state..." of DNA sequencers;
the 373A, with the Macintosh environment and different software is a much
different instrument than the old 370A.  If the people who are complaining
about the 370's user unfriendliness, etc. (and there are many!) would simply
buy a Mac and upgrade their instruments, they would see a world of difference.
Secondly, and perhaps more critical to successes with the automatic DNA
sequencers, the chemistry has greatly improved.  T7 DNA polymerase chemistry
and the Taq linear amplification sequencing method have made the 373A and
Pharmacia's ALF extremely useful and easy to use.  We routinely sequence from
c. 400 ng of double-stranded plasmid DNA, prepared using a standard mini-prep
procedure.  The sequencer usually gives good reads (99%+ accuracy) of 470-500
base pairs, and we typically analyze 48 samples per instrument per day.  I've
run a lot of gels (and read a lot of films) and I've got to tell you that I've
gone fluorescent and I'll never go back.

As an example of how easy the 373A is to run these days, I have an undergrad in
lab who didn't know a dideoxynucleotide from a diode two weeks ago.  Now, she
is running one of our 373s everyday (24 templates per run), and will probably
finish sequencing one or two cosmids before the end of the summer program.  So
much for the need for a "skilled technician".

For those who are contemplating purchase of an automated DNA sequencer, you have
some choices.  The ABI 373A is easy to use, is compatible with T7 (Sequenase) or
Taq sequencing methods, allows high-throughput sequencing with good accuracy, bu
t is currently limited to dye-primer sequencing.  Furthermore, since the 373A
uses 4-color chemistry, synthesis and use of custom dye-primers is tricky.  The
ALF gets around this problem with its one-color dye-primer chemistry, and custom
primers may be labeled with fluorescein amidite as part of primer synthesis.
The ALF is lower throughput than the ABI (10 samples per run), but accuracy
appears to be roughly the same over 500 base pairs.  Software on the ALF is
still a little rough, but it's a new machine, and I expect the software to
improve significantly with the next release (I've seen a beta version).  Cost
of the two machines is about the same; reagent cost will vary depending on the
chemistry you choose.

As far as software for editing, assembling, analyzing and managing sequencing
data, it is quite simple to send all of the raw data files (not sequence text
files) over an network connection to another computer.  A second Mac can be
used with the ABI software to view, edit and print data; a Sun workstation
(a SPARCstation IPC, cost about $7000) can be set up with Rodger Staden's
new Xwindows programs for viewing, editing and assembling data from either the
373A or ALF sequencers.  Alternatively, a Mac IIci with 8 MB of RAM running
Multifinder seems to do just fine (thouhg a little slow) with other tasks while
collecting data from the 373A.

So, my advice to those considering fluorescent sequencers would be very positive.
We have three of 'em, one ALF and two ABI, and good DNA sequence data is going
into the computer faster than with any other method I've ever used.  If you
want some more info, give me a shout.

Rick Wilson, St. Louis
rick at
Disclaimer:  I don't work for ABI, Pharmacia, Sun Microsystems or Rodger Staden;
                I just sequence a lot of worm DNA.

More information about the Methods mailing list