Hi Oded, I think the old 5.2.0 might be under estimating the error rate as compared to the new release candidate. You cannot see the decoys here because of the settings you used. However your database has two independent sets of decoys available DECOY0 and DECOY1. Can you use one of the set for PeptideProphet and use the other set to get a decoy count at a given PeptideProphet probability cutoff. I suspect the 5.2.0 TPP will show more unknown decoys than the new release candidate. Can you test if this is the case here? I can run a more in depth analysis when I am back from vacation next week.
Cheers, David On Tue, Apr 13, 2021, 2:07 PM Oded <oded.kleif...@gmail.com> wrote: > Dear David, > This is PeptideProphet issue (we don't use iProphet for this analysis). > With the search parameters that I sent you followed by running the > PeptideProphet command of *xinteract -Ninteract.pep.xml -p0.05 -l7 -PPM > -OANEp -dDECOY Seq69478_QE2.pep.xml,* > and lastly using a cutoff of 1% error rate we obtain with TPP 5.2 2836 > correct assignments and with TPP 6.0 only 2228 correct assignments. The > probability for 1% error rate is 0.84 in TPP 5.2 and 0.9090 in TPP 6.0). > In the following link, you will find the interact files (new and old): > https://www.dropbox.com/t/mwtpCLeEJRLCwbi1 > <https://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.dropbox.com%2Fl%2FAACfc2b_KukHZAS1VuNo4sPM3gEXyXtDlPE&data=04%7C01%7Coklab%40technion.ac.il%7C342d2db781c2443416b108d8febfbc70%7Cf1502c4cee2e411c9715c855f6753b84%7C1%7C0%7C637539447557029368%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C2000&sdata=LBdAB3QsH9rUxvNKMwKjOKYDbQXFxma9dREKgS5siWQ%3D&reserved=0> > Thanks, > Oded > > > > On Tuesday, 13 April 2021 at 03:38:33 UTC+3 David Shteynberg wrote: > >> Dear Oded, >> >> Thanks for this. I ran a quick test and I actually observed a few more >> PSMs for PeptideProphet 6.0.0-rc14 for the same PeptideProphet probability >> cutoff for this dataset. Is the issue you see with PeptideProphet or >> iProphet results? Which spectra were getting excluded in the your testing >> of 5.2.0 vs 6.0.0, can you give me some specific examples? Are you certain >> you are using identical analysis parameters for PeptideProphet and what are >> those parameters? >> >> Cheers, >> >> David >> >> On Mon, Apr 12, 2021, 2:15 PM Oded <oded.k...@gmail.com> wrote: >> >>> And also the search results with the new vs the old version (although I >>> think they are more or less the same): >>> https://www.dropbox.com/t/8VyBVdeuoWdk982q >>> >>> >>> On Tuesday, 13 April 2021 at 00:09:05 UTC+3 Oded wrote: >>> >>>> Hi David, >>>> You can find the mzML, Comet parameters and FASTA file here: >>>> https://www.dropbox.com/t/Yld0ZBWhsuFajeLj >>>> The link is valid for 7 days. >>>> Many thanks, >>>> Oded >>>> >>>> On Monday, 12 April 2021 at 23:46:08 UTC+3 David Shteynberg wrote: >>>> >>>>> I don't mind taking a crack at it over vacation. Please let me know >>>>> where I can pull the data from. I might not have quick solution for you >>>>> but >>>>> I can get started looking for the problem. Which search engine did you >>>>> use >>>>> here? I would need mzML data search results and the fasta database to >>>>> get >>>>> started. >>>>> >>>>> Thanks! >>>>> >>>>> >>>>> >>>>> Thanks! >>>>> >>>>> On Mon, Apr 12, 2021, 1:38 PM Oded <oded.k...@gmail.com> wrote: >>>>> >>>>>> Hi David, >>>>>> Thank you for the quick reply. We are using TPP v6.0.0-rc14 >>>>>> Noctilucent, Build 202103031119-8400 (Windows_NT-x86_64). >>>>>> As for the data let me know when you are back and I will transfer you >>>>>> either the raw file or the search results. >>>>>> Enjoy your vacation. >>>>>> Thanks, >>>>>> Oded >>>>>> On Monday, 12 April 2021 at 22:17:53 UTC+3 David Shteynberg wrote: >>>>>> >>>>>>> Hi Oded, Which release candidate are you referring to? The earlier >>>>>>> candidates may have a bug that is corrected in a later version. If you >>>>>>> can >>>>>>> share some data and specifics about the missing PSMs I can run it here >>>>>>> and >>>>>>> troubleshoot the problem. >>>>>>> >>>>>>> Thanks! >>>>>>> >>>>>>> David >>>>>>> >>>>>>> P.S. I am on vacation this week so will troubleshoot next week. >>>>>>> >>>>>>> On Mon, Apr 12, 2021, 11:49 AM Oded <oded.k...@gmail.com> wrote: >>>>>>> >>>>>>>> Hi there, >>>>>>>> We recently download the TPP 6.0 RC and while using it we noticed >>>>>>>> that we obtain fewer peptides IDs than what we got for the same >>>>>>>> dataset and >>>>>>>> search output with version 5.2. >>>>>>>> Many of the missing peptides seem to have decent Expect value and >>>>>>>> MS/MS following a visual inspection. >>>>>>>> This seems that this is due to changes in PeptideProphet and are >>>>>>>> not related to the search itself that was done with Comet. >>>>>>>> Is there any additional parameter that should be included in the >>>>>>>> new version in order to restore the missing peptides? >>>>>>>> Many thanks, >>>>>>>> Oded >>>>>>>> >>>>>>>> -- >>>>>>>> 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-discu...@googlegroups.com. >>>>>>>> To view this discussion on the web visit >>>>>>>> https://groups.google.com/d/msgid/spctools-discuss/b7c2e3b7-3061-4ec7-94d8-f70e43538f2an%40googlegroups.com >>>>>>>> <https://groups.google.com/d/msgid/spctools-discuss/b7c2e3b7-3061-4ec7-94d8-f70e43538f2an%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-discu...@googlegroups.com. >>>>>> >>>>> To view this discussion on the web visit >>>>>> https://groups.google.com/d/msgid/spctools-discuss/7f5e47e5-56df-4552-88ca-b6d4396f3d8en%40googlegroups.com >>>>>> <https://groups.google.com/d/msgid/spctools-discuss/7f5e47e5-56df-4552-88ca-b6d4396f3d8en%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-discu...@googlegroups.com. >>> >> To view this discussion on the web visit >>> https://groups.google.com/d/msgid/spctools-discuss/3baf2dd3-f237-4c9d-82a2-54d563ccf5e6n%40googlegroups.com >>> <https://groups.google.com/d/msgid/spctools-discuss/3baf2dd3-f237-4c9d-82a2-54d563ccf5e6n%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/3030a2db-c902-4cb4-837d-20f385662a18n%40googlegroups.com > <https://groups.google.com/d/msgid/spctools-discuss/3030a2db-c902-4cb4-837d-20f385662a18n%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%3D80-7ocyCdCPUPbSP%2BeJCqVWcvYjQiwsptMzsMJ5yM2Pw%40mail.gmail.com.
