Jeff, TPP 4.1 is really old (2008) and the Sashimi SourceForge logs only go back to 09/2009. There have been a lot of changes and bug fixes in the past 5 years, some minor and others more significant. If you really are curious, your best bet to see a list of changes would be to look at the commit messages for XPressPeptideParser.cxx which you can find here: http://bit.ly/19DypHd
Here are things to try with your data: 1. Use appropriate mass tolerance to reconstruct chromatograms; don't use a tolerance that's too narrow. There's a ppm option if that's pertinent for your data. 2. Consider using a fixed scan range from the peak apex (-F option on command line); this is your only control to limit the elution peak scan range. 3. If you're using high res MS scans with a narrow mass tolerance, have XPRESS use one or more isotope peaks in additional to the monoisotopic peak for chromatogram reconstruction (-p option) Good luck. I hear good things about MassChroQ so consider using that tool to do your quantitation. Or if you want to stay within the TPP, you could always try ASAPRatio to see if that gives you better results. - Jimmy On Mon, Dec 9, 2013 at 1:35 PM, Jeffrey Milloy <[email protected]>wrote: > Hi all, > > I am comparing XPress peptide results between TPP versions *4.1* and *4.6*. > Given the same input pep.xml file and mzXML file, the results from the two > versions are quite different (R-squared of ~0.3). To my surprise, the 4.1 > results seem much better. Let me provide some details, and maybe someone > can help me improve my results. > > > *Version and Parameters* > > TPP 4.1 > <xpressratio_summary version="2.1 (TPP v4.1 JETSTREAM (RC) rev 0, Build > 200806181121 (linux))" author="Jimmy Eng" same_scan_range="Y" > labeled_residues="Kn" xpress_light="1" massdiff="K,8.04 n,8.04" > masstol="0.02"/> > > TPP 4.6 > <xpressratio_summary version="2.1 (TPP v4.6 OCCUPY rev 3, Build > 201308191103 (linux))" author="Jimmy Eng" same_scan_range="Y" > labeled_residues="Kn" xpress_light="1" massdiff="K,8.044370 n,8.044370" > masstol="0.020000" ppmtol="0" min_num_chromatogram_points="6" > min_num_isotope_peaks="0"/> > > > *Specific Result* > > *pep.xml input:* > > scan: 1576 > precursor_neutral_mass: 2492.1902 > assumed_charge: 3 > retention_time_sec: 1771.0 > peptide: GQKSPGALETPSAAGSQGNTASQGK > calc_neutral_pep_mass: 2492.1908 > massdiff: -0.000630 > mod_nterm_mass: 29.039125 > mod_aminoacid_mass > - position 3, mass 156.126263 > - position 4, mass 166.998358 > - position 25, mass 156.126263 > > *TPP 4.1 xpressratio result:* > > light: scans 1555-1604, mass 2493.199, area 3.13e+06 > heavy: scans 1555-1604, mass 2517.332, area: 3.43e+06 > decimal_ratio: 0.91 > > *TPP 4.6 xpressratio result:* > > light: scans 1524-1649, mass 2493.1981, area 3.19e+06 > heavy: scans 1524-1644, mass 2517.3312, area: 3.43e+06 > decimal_ratio: 0.93 > > *Notes:* > > - The 4.1 light mass is 1.008 Da greater than the peptide neutral > mass, and the 4.6 light mass is 1.0073 Da greater. In this regard, I > believe *version 4.6 is correct*, because the light mass equals the > neutral mass plus the mass of a proton (rather than the mass of hydrogen). > > > - The 4.6 result has a very wide scan window. In this regard, the *4.1 > result seems correct*. The image below is an extracted ion > chromatogram, with the 4.6 result width. On the other hand, the 4.1 result > starts and stops where you would expect, around 19.25s - 29.75s. > > > > <https://lh6.googleusercontent.com/-kyql1liKr5k/UqJXaoatn0I/AAAAAAAAAIQ/Yysv4stwVJA/s1600/1576.xic.png> > > > *Comparison* > > The first scatter plot (n=197) shows log2 ratios for the two versions, > with a poor correlation. The second two plots show the xpress ratios vs > MasschroQ ratios for confirmation. You can see that the old TPP 4.1 results > correlate much better with MassChroQ. > > > > <https://lh4.googleusercontent.com/-ak_V4WFznps/UqYxM0Iu-LI/AAAAAAAAAIg/h4TFkCe_uOw/s1600/xpress_oldvnew.png> > > > <https://lh3.googleusercontent.com/-JCh0394YMf8/UqYxTpJa2mI/AAAAAAAAAIo/td4xnbyYAno/s1600/xpress_mcqvold.png> > > > <https://lh5.googleusercontent.com/-mH9zEfFJwf0/UqYxaHxSbLI/AAAAAAAAAIw/7RjADenxX2E/s1600/xpress_mcqvnew.png> > > > *Summary* > > There seem to be changes in xpress between TPP 4.1 and 4.6, despite the > version number remaining the same, which result in significantly different > quantification. > > - Is there a changelog for xpress? > - Am I correct about light and heavy mass calculations? When did those > change? > - How can I improve the peak start/end scans in the current (TPP 4.6) > version of xpress so that they are not so wide? > - Has any one else found that xpress quantifications have become > seemingly *less* accurate in recent versions? Are there important new > parameters that I need to tweak, or some other change that is necessary > with the new version to improve my results? > - If I were to revert to a previous version of xpress, should I revert > all the way to 4.1, or is there an intermediate version that would be > preferable? > > > Thanks so much, > and let me know if there is anything else useful that I can provide > > Jeff > > -- > You received this message because you are subscribed to the Google Groups > "spctools-discuss" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To post to this group, send email to [email protected]. > Visit this group at http://groups.google.com/group/spctools-discuss. > For more options, visit https://groups.google.com/groups/opt_out. > -- You received this message because you are subscribed to the Google Groups "spctools-discuss" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/spctools-discuss. For more options, visit https://groups.google.com/groups/opt_out.
