A bit late for a response, but I was also experimenting with QualScore and encountered the same issue, where I'm also getting NaN coefficients and 0 dtas. After debugging the QualScore source code from here <https://sourceforge.net/p/sashimi/code/HEAD/tree/trunk/qualscore/>, the issue seems to be with their custom mzXML parser class (org.isb.mzxml.MzXMLParser.java).
It looks like the bug occurs when the code tries to parse the encoded mz_intensity list of a scan from the MzXML file. In the function MzXMLParser.convertMassList(), the code assumes that the mz and intensity values are encoded in a 32-bit single-precision binary floating-point format, which would work fine if that's the format of your MzXML file. However, in my raw file at least, the mz_intensity values are encoded in 64-bit precision doubles, so the incorrect parsing ends up converting the mz and intensities to very small values, which is probably what led to the NaN coeffs and 0 dtas further down the line. I couldn't figure out how to convert my mzXML file such that it encodes mz and intensity in 32-bit precisions instead of 64-bit (even the --mz32 --inten32 options from ProteoWizard's msconvert don't seem to do the job), which is probably the easier way to solve this issue. So, I ended up modifying the source code to make it parse the file as 64 bits precision doubles instead, as shown in the screenshot below, which solved the NaN and 0 dta issue for me. The modification was made in the convertMassList() function and the endElement() function, both under the MzXMLParser.java class mentioned above. [image: CodeModification.PNG] Hope this might help. Best regards, Yun-En Chung On Thursday, July 16, 2020 at 2:10:50 PM UTC-4 migue...@gmail.com wrote: > Dear all, > > I am very interesting on using QualScore. > > I managed to get it running within the TPP GUI but I believe there's a > problem with the output. It seems the tool detects the unassigned spectra > but none is then written in the output. The qdir.coef file contains only > NaN values. > > I am including here the command I used along with the output log. > > *EXECUTING: cd c:/TPP/data/testing_qualscore && c: && java -jar > C:/TPP/bin/qualscore.jar -c 1.0 -p 0.1 > hek_2_test1_p0_settrue_interact.pep.xml* > > > QualScore version 1.0_2 Copyright (C) 2006 > Institute for Systems Biology, Seattle WA > QualScore is free software and comes with absolutely no warranty. > > XML > Loading interact data: > 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 > 12000 13000 14000 15000 16000 17000 18000 19000 20000 21000 > 22000 23000 23145 > Interact data file contains 23144 entries. > 0 +1 were removed. > Enumerating scans: > processing HEK_2.mzXML... 500 1000 1500 2000 2500 3000 3500 4000 > 4500 5000 5500 6000 6500 7000 7500 8000 8500 9000 9500 10000 10500 11000 > 11500 12000 12500 13000 13500 14000 14500 15000 15500 16000 16500 17000 > 17500 18000 18500 19000 19500 20000 20500 21000 21500 22000 22500 23000 > 23144 > Features calculated for: 23144 (15082 ms) > Training set: 8909/8666 (good/bad) > Unassigned spectra: 11165 > Quality score calculation time: 74 ms > Writing classifier co-efficients to file: > hek_2_test1_p0_settrue_interact.pep.qdir.coef > *0 dtas written to* hek_2_test1_p0_settrue_interact.pep.qdir. > Time: 15 > > I am running this on Windows 10. I generated the interact.pep.xml setting > up the command -p0 in peptideProphet to include all assignments. The > pep.xml file consist of a search result from Comet, executed from the same > instance of TPP. > > Does anyone has an idea why this could be happening? I am having similar > results with two different raw files from different measurements runs. > > In case this tool is not supported anymore, does anyone has a suggestion > of other tools or approach to salvage high-quality but unassigned spectra > after DB search? > > I would really appreciate any help. > > Best regards, > Miguel > > > -- 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 spctools-discuss+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/spctools-discuss/18919b46-4489-44e0-abae-49fb763671f7n%40googlegroups.com.
