Thankyou for explaining what is happening in this case. 1. I further want to confirm which model I should prefer as my objective is to identify the maximum number of proteins present in the sample. 2. Also how to incorporate decoys in the database. Regards Reeju
On Mon, Sep 2, 2024 at 8:52 PM David Shteynberg < [email protected]> wrote: > Hello Reeju, > > Thank you for trying the TPP for you analysis and for your questions and > slides. Here is what I think might be happening. In 6.3 the scores from > comet and the PeptideProphet analysis based on the models were different > from 7.1. > > Here is the comet f-value model in 6.3: > > [image: PastedGraphic-1.png] > > Here is the f-value model from 7.1: > > [image: image.png] > > > The goal of PeptideProphet is to identify a bimodal distribution and fit > two distributions to the observed black line, with the corrects modeled in > green and the incorrect in red. In the first case (6.3) the observed > distributions is not really bimodal to the eye and the PeptideProphet model > is fitting both distributions to the one peak because the observed > distribution is not bimodal enough, and please correct me if I am wrong, > you are not providing known true negatives (decoys) to help train the model > and keep the red distribution “pinned" by them. In the second case (7.1), > the bimodal distribution is fitting much better in my opinion as the red > curve is mostly in the correct place and the green model is fitting the > shoulder where the majority of the correct PSMs in your data will be. I > know you feel like you “lost” some IDs going fro m 6.3 analysis to 7.1, but > I would say you likely improved the confidence in the results by the > reanalysis. You can verify this by incorporating unknown true negatives > (entrapement decoys) in you database to use as another FDR estimate to > compare to the model’s results from 6.3 and 7.1. > > > Cheers! > -David > > On Sep 2, 2024, at 5:44 AM, reeju <[email protected]> wrote: > > Hello everyone, > I am running a raw file in TPP 7.1, which was also analyzed previously in > an earlier version via TPP 6.3. I have kept the comet parameters and > peptide prophet settings similar for both the cases, still there is > difference in proteins identified from both TPP versions. In tpp6.3 the > total proteins identified were 1546, after applying filters on the same > file with probability 0.88(having 0.05 error rate) the protein count was > still 494. But in tpp 7.1 the total protein count displayed was only 99 > without any filtering on the same file. Why is there such a drastic > difference. Comparison of both files have been attached. If anyone can help > us troubleshoot please revert back. > Thank You > Regards > > > -- > Reeju Rani > Research Scholar > Proteomics and Cell Biology lab > ICAR-National Dairy Research Institute > Karnal, Haryana India > > -- > 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 view this discussion on the web visit > https://groups.google.com/d/msgid/spctools-discuss/CAJKhLXzf3UGPXKrergg-s5VuOtciS2O2Vq18tgu2WncQcB5L4g%40mail.gmail.com > <https://groups.google.com/d/msgid/spctools-discuss/CAJKhLXzf3UGPXKrergg-s5VuOtciS2O2Vq18tgu2WncQcB5L4g%40mail.gmail.com?utm_medium=email&utm_source=footer> > . > <comet.params><comet.params><TPP.pptx> > > > -- > 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 view this discussion on the web visit > https://groups.google.com/d/msgid/spctools-discuss/FCA9687F-9C6B-45E5-A917-4B400C05F189%40systemsbiology.org > <https://groups.google.com/d/msgid/spctools-discuss/FCA9687F-9C6B-45E5-A917-4B400C05F189%40systemsbiology.org?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 [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/spctools-discuss/CAJKhLXyTqZaXEJW0bFbJ%2BcEx%2B8aEpbAjE7MV4Ttoa3x21hfZyw%40mail.gmail.com.
