It would be very helpful to see your Kojak pepxml results. Thanks, David
On Fri, Sep 13, 2019, 12:01 AM Thomas Gossenreiter < thomas.gossenrei...@gmail.com> wrote: > Hi David, > > thanks for your quick response. Here is the link to the zipped pep.xml > files. Is this enough or do you also need the pep.xmls from the kojak > search? > > https://drive.google.com/open?id=1zKW0JVfVL_LME0gmt_L6AI7j8AymUmD1 > > > Am Donnerstag, 12. September 2019 18:02:07 UTC+2 schrieb David Shteynberg: >> >> Hi Thomas, >> >> I am not sure what is happening but if you could provide the dataset >> files for the two search results I can investigate. PeptideProphet does >> calculate different models for the different crosslink types. >> >> Thanks, >> -David >> >> On Thu, Sep 12, 2019 at 8:59 AM Thomas Gossenreiter < >> thomas.go...@gmail.com> wrote: >> >>> Hi, >>> >>> I have used Kojak 1.6.1 in the TPP pipeline and got some strange results >>> (very similar ms2 spectra from one peptide with similar Kojak scores got >>> very different probabilities when running peptide prophet). I have plotted >>> the Kojak score vs. the peptide prophet probability for target and decoy >>> hits and there is hardly any correlation between the kojak score and the >>> probability. >>> >>> I have repeated the search Kojak version that comes along with TPP >>> (1.5.5) using the same settings except for the top_count parameter. I think >>> this looks how it should, except for the cross-links which are scored very >>> well by Kojak but get a bad peptide prophet probability. Is this the case >>> when one peptide has a very good Kojak sub score and the other one a very >>> bad score? >>> >>> Does version 1.6.1 work for you or do you have any idea what is >>> happening here? I attached a ppt file with the figures. >>> >>> I also got a general questions: Is peptide prophet calculating FDRs >>> separately for cross-linked, loop-linked and normal peptides? >>> >>> Kind regards, >>> Thomas >>> >>> >>> >>> >>> -- >>> 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...@googlegroups.com. >>> To view this discussion on the web visit >>> https://groups.google.com/d/msgid/spctools-discuss/12f12195-f115-41e5-a196-b53e18e170df%40googlegroups.com >>> <https://groups.google.com/d/msgid/spctools-discuss/12f12195-f115-41e5-a196-b53e18e170df%40googlegroups.com?utm_medium=email&utm_source=footer> >>> . >>> >> -- > 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/abe43fc7-eb03-4411-a6ac-c5ca4a5c326a%40googlegroups.com > <https://groups.google.com/d/msgid/spctools-discuss/abe43fc7-eb03-4411-a6ac-c5ca4a5c326a%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- 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/CAGJJY%3D-X2sxPCq75ryeX0%2BAO6rGodAKtJ%2BoDVzePQM%2BbkLPmDQ%40mail.gmail.com.