Bruker Data Analysis .mgf files into the TPP?

[Chris Rath]

Hi everybody:

I have a large proteomics LC-MS/MS dataset that I've generated on a
Bruker Q-FTICR MS, and would like to run it through the TPP as I am
successfully doing with my Orbitrap datasets.

I know I have a number of high quality hits (X!tandem expect values
1e-7, Mascot hits, same peptide in Orbi dataset, match with standard
peptides) but I am either not getting data processed through TPP, or
getting odd results--likely due to file conversion issues. For
example:

1. Compassxport in line spectra mode (.baf->.mzxml) results in poor
quality spectra and no IDs.

2. Compassxport in profile spectra mode (.baf ->.mzxml) crashes on my
10-20Gb input files.

3. Data Analysis export of .mgf files followed by X!Tandem followed by
pep.xml conversion results in pep.xml files with visible X!tandem
hits, but no probabilities produced by peptideprophet >0. Is there a
reason peptide prophet cannot assign probability scores to pep.xml
files generated from .mgf input?

4. Data Analysis export of .mgf conversion, perl script to .dta, perl
script .dta to mzxml results in TPP results, however my final .mzxml
files are scrambled with incorrect scan numbers.

Obviously this is a complicated issue but any insight would be
appreciated.

Alternatively, does anyone know researchers running Bruker datasets in
the TPP whom I could contact directly?.

Thanks,

-Chris

[A Hill]

Chris,

FWIW, we took a similar file format conversion path for Bruker data,
but unfortunately I don't have experience with the X!Tandem/TPP
issues.

The processing path we took was:

- Data Analysis export of XML file (centroided "line" MS2 spectra),
followed by perl script conversion to .dta (the XML format contains
some metadata like compound indexes for each scan that was useful for
downstream processing).
- The .dta files were submitted to Sequest and Mascot, giving the
expected search results.

For your #4, do you know if the .dta to mzXML converter you are using
expect a specific .dta file name format like <sample>.<first
scan>.<last scan>.<charge>.dta? If so, and your .dta file names are
different than this format, perhaps that is contributing to the
scrambled/incorrect scan numbers in the generated mzZML.

Andrew.