It has been my experience that ASAPRatio is slow (haven't used 4.8 yet) and inaccurate even with high mass accuracy data. I've used both chemical isotopic labeling and SILAC data sets and you really need to QC the results. The output of ASAPRatio is not to be trusted, based on my experience with it. I've seen "interesting" hits have a given "interesting" ratio and then go back and look at the peak integrations and have occasionally found that after performing QC on the data that the ratio was reversed from what was originally reported by the algorithm! This calls into question the accuracy of the entire dataset quantitation as performed by ASAPRatio.
The original Aebersold paper on ASAPRatio is demonstrated on a reduced complexity dataset e.g. a pulldown. The P-value was described as a way to quickly sort changed from unchanged compared to the background. I am suspicious that ASAPRatio is just not that suited to datasets from complex samples like cell lysates et. al. which is what I have been using it for and I suspect you have as well. I will begin use of XPRESS and see how that performs. Given the need for quantification of proteomic data sets, I hope the TPP team is working on a better MS1 quan package than what is currently present in TPP. It would be nice to see something more robust like Skyline but better suited to shotgun vs targeted quantification. There is a paper describing ISOQUANT which looks promising but I believe it was originally written to handle SILAC only datasets. Maybe the TPP developers can obtain this from the author and adapt it to handle multiple isotope labeling strategies and make it available to all their loyal TPP users who have benefited greatly by their efforts to date. Since you have high resolution data, have you tried MaxQuant? Cheers, Brian Brian Hampton Protein Analysis Lab Center for Vascular and Inflammatory Diseases University of Maryland School of Medicine 655 West Baltimore Street BRB 7-018 Baltimore MD 21201 V: 410-706-8207 On Mon, Dec 1, 2014 at 1:04 PM, Alejandro <[email protected]> wrote: > A quick update. > > After running the same files with TPP 4.8, running time with ASAP ratio is > quite reduced, to ~1h, but quantitation seems to work better with XPRESS. > > Alejandro > > On Wednesday, November 19, 2014 7:29:30 PM UTC+1, Alejandro wrote: >> >> Hi all, >> >> I'm analyzing dimethylated data obtained with a QExactive using XTandem! >> (have tried both static or variable searches). If the data is quantified >> using Xpress, the time it takes is quite normal, ie fast, minute(s) or so >> for a 1GB RAW file (ca. 250MB mzXML), however when activating ASAPratio, >> the processing takes hundred times more than with Xpress (which reports a >> quite disperse distribution of ratios, not seen with Orbitrap data, thus >> the reason to see how ASAP performs). After looking in the list I found a >> post which discussed somehow this https://groups.google. >> com/d/topic/spctools-discuss/TkkoHpXAh4A/discussion which was related to >> multithreading. >> >> When running this analysis the bottleneck seems to come from the >> ASAPratio: >> >> Stucked at: >> >> running: "C:/Inetpub/tpp-bin/XPressPeptideParser "interactasap.pep.xml" -m20 >> -a -nn,8.04437027 -nK,8.04437027 -H -c5" >> .................................................. 1k >> .................................................. 2k >> .................................................. 3k >> .................................................. 4k >> .................................................. 5k >> .................................................. 6k >> .................................................. 7k >> . >> command completed in 22 sec >> >> running: "C:/Inetpub/tpp-bin/ASAPRatioPeptideParser "interactasap.pep.xml" >> -lnK -C -r0.02" >> >> >> The same scenario as the one described in the old post seems to happen >> here, only one core and ~50MB usage of RAM, using only >> ASAPRatioPeptideParser.exe and this taking several hours to complete. >> >> Could it be also the mentioned "...redundant disk reads and base64 >> decoding due to poor caching of the scans read from mzXML..." the cause? Or >> it is just the nature of the data obtained with the Qexactive that makes it >> take longer? I haven't checked if something would change by using an old >> version of ASAPRatio ~2009 from that post. Though I don't think this would >> be compatible with the current parsers. >> >> Any clues? >> >> Thanks, >> >> Alejandro >> > -- > 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/d/optout. > -- 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/d/optout.
