Staden (with phred/phrap) vs. Consed (with phred/phrap)

[James Bonfield]

In article <388FBFE5.1065EE90 at mail.ru> Volodya <bo-ba at mail.ru> writes:
>	Thank you very much for your input. What do you think about the use of
>phred scores in Staden and Consed? I am interested in using that package
>that allows fuller usage of this objective quality measure, and possibly
>a lesser human intervention and better integration within the automated
>system. Are Consed and Staden equivalent in this regard, if they are set
>up to use phred and phrap? On the other note, is it too much to ask you
>to send me your perl script? It would be very useful for a quick test
>drive of Staden/phred/phrap package, since I hear that setting it might
>be a problem.

Hello Volodya,

I implemented the use of phred scores within Gap4 so I can explain how
this works, although I think that the best source of information for
Gap4 is in the following web pages:

http://www.mrc-lmb.cam.ac.uk/pubseq/manual/gap4_121.html#SEC194
http://www.mrc-lmb.cam.ac.uk/pubseq/manual/gap4_122.html
http://www.mrc-lmb.cam.ac.uk/pubseq/manual/gap4_55.html#SEC85

For what it's worth, the Sanger Centre are now routinely using Gap4 in
the "confidence" consensus mode to direct their final editing. The
original estimates showed that this final editing stage was sped up
around 7 fold (compared to their original strategy of checking all
discrepancies not automatically fixed by auto-edit) with comparable
final consensus accuracy.

I cannot well explain how this compares to Consed. I have only seen
consed once, and I didn't use it myself. Indeed we have deliberately
avoided asking for a copy of Consed to ensure that our ideas are fresh
(although sometimes the solution is obvious). However from what I gather
consed cannot calculate a consensus itself. Rather it uses the
consensus given to it by phrap. Hence the confidence values assigned
to consensus bases are those that phrap determines. Unfortunately we
can no longer obtain any newer versions of phrap, so I do not know how
well the newest versions compare to the release we have (0.960731),
but I believe the phrap consensus algorithm has been recently changed.

	James
--
James Bonfield (jkb at mrc-lmb.cam.ac.uk)   Tel: 01223 402499   Fax: 01223 213556
Medical Research Council - Laboratory of Molecular Biology,
Hills Road, Cambridge, CB2 2QH, England.
Also see Staden Package WWW site at http://www.mrc-lmb.cam.ac.uk/pubseq/