\\mene\bioinfo\Proteomics\Kuerer\KRC-WCX2-24QC\readme.txt

17-18 December 2003
Kevin R. Coombes

1. The XPT files from 24 identical samples (run on 3 chips, two spots at a
   time, over a four day period) were loaded into a single Ciphergen
   experiment file in version 3.2.0.094 of the Ciphergen software.
   Note that spots A and B of each chip were used on the first day, spots C
   and D on the second day, etc.

2. The spectra were filtered, baseline-corrected, and normalized using the
   default settings in the Ciphergen software.

3. Peaks were detected using the default settings, and a peak list was
   exported to the file "kuererqcPeaksDefault.csv". Using this setting, we
   found a total of 224 peaks, which averages at about 9 peaks per
   spectrum.

4. Peak finding sensitivity was reset to "maximum" using the
   slider. Existing peak labels were cleared and the peak finding was
   repeated. Results were exported to the file "kuererqcPeaksMaxSens.csv".
   Using this setting, we detected a total of 996 peaks, which averages out
   to 41.5 peaks per spectrum.

5. All unprocessed spectra were then exported in XML format to the
   subdirectory "RawXML" of the current directory. As expected, this
   procedure yielded 24 distinct XML files. 

6. The XML files were converted to our two-column raw text format using the
   script "xml2txt.pl" in the AnalysisScripts subdirectory. Note that this
   script relies on the "ciphergenXMLreader.pl" script in the directory
   /Proteomics/bin as of 17 December 2003. Note also that the script
   truncates the file name to a unique identifier of the form W###A, where
   the W is literal (and identifies the WCX2 chip), the three digits
   identify the exact chip, and the final letter identifies the spot within
   the chip. The script also verifies the existence of the letters QC in
   the original file name, corresponding to an actual instance of the QC
   sample.

7. A listing of the raw spectra was created using the command:
      fnames.pl RawSpectra RawFiles.txt

8. A matrix combining all the raw spectral data was prepared in MATLAB
   using the function "collateSpectra.m" that is contained in the
   Cromwell package. This matrix was saved in the file "rawSpec.mat"
   in the main directory. We also saved the mass values (which were
   the same for all files) in the file "rawMass.mat".

9. After creating the rawSpec.mat and rawMass.mat files, the text versions
   in the RawSpectra directory were deleted and the XML versions in the
   RawXML directory were placed into a zip file. (Care may need to be taken
   when unzipping on a UNIX machine since the zip file was created on a
   Windows machine with the default settings. This can lead to differences
   in end-of-line characters across operating systems.)

10. Record keeping now passes to the MATLAB file "collate.m", which
    performs the main smoothing.