bug in ms-sequest 0.0.5

[Gabe]

Hi John,

The 'srf_to_sqt.rb' script worked fine in 0.0.4, but when I upgrade to
ms-sequest-0.0.5 I get a "bad peptide sequence: °╦☺" error and it
crashes out immediately, without generating an SQT file. When I
downgrade to 0.0.4 the error goes away and the conversion completes
successfully.

FYI - in both versions (including 0.0.4 which works fine) I get a
"Band04.srf has no SEQUEST information, may be a truncated/corrupt
file!" error.

Thanks

Gabe

[John Prince]

Your files revealed a bug in my understanding of the Bioworks srf file.  (one might describe it as a bug in the way the srf file works)

The gory details:
In the srf file, Bioworks respects print_duplicate_references if it is 0, but not for values greater than zero (e.g., if print_duplicate_references = 0, then no extra proteins are listed in the srf file.  If print_duplicate_references is > 0, the value is not respected, so that the srf file will list, say, all 20 proteins even though the value is set at 10.

My test files don't run into this problem, but yours did.  version 0.0.7 fixes this and I can transform your Band4.srf file, too.

gem install ms-sequest
gem list --local ms-sequest  # should be version 0.0.7, if not wait until the mirror has it (usually < 1hr)

--John

[gabe simon]

John,

First, this is of minor/no-importance to me: ms-sequest v0.0.4 works
fine for my purposes (I think) so please don't worry about this if
you're busy with other things. That said, v0.0.7 doesn't work here.
I removed all old versions (gem uninstall ms-sequest -a) and installed
v0.0.7 and now I get an additional error:

Band04.srf has no SEQUEST information, may be a truncated/corrupt

file! [I always get this error]
no params file found in srf, could be truncated file! [this is
the new one]
bad peptide sequence: °╦☺ [i got this in v0.0.5, as well]

What you said about the print_duplicate_references sounds like
Bioworks is treating this value as a boolean: 0 = false, >0 = true,
rather than a number of duplicates to include. I'm not sure if that
has anything to do with this new error, though.

Best,

Gabe

[John Prince]

Gabe,

I really do appreciate you letting us know as the whole point of an open, public library is that it works for the public :)  Its no real trouble--I'm just anxious that it perform properly.

OK, I fixed the error.  [Forgot to open the srf file in windows binary mode when scanning for the sequest.params file]. 

Version 0.0.8 converts Band04.srf on windows (NT anyway).

gem install ms-sequest
gem list --local ms-sequest

[needs to be 0.0.8] 

--John

[gabe simon]

excellent. no error messages at all! :) I'll keep you posted as to
how it works for my collaborators. Thanks again!

best,

gabe

Thanks again for everything. FYI - I've