|
Dear Skyline Users, People were asking questions during our fourth Skyline Tutorial Webinar on targeted method design so ... here are the answers! Q: Is it possible to calculate the retention time of a peptide you haven't measured by now depending on the retention time information of a few spiked in peptides for a specific gradient? (like BSA for example) Ans: If you have no prior measurement of your peptides at all, then you can still use the SSRCalc (Sequence Specific Retention Calculator - http://www.ncbi.nlm.nih.gov/pubmed/20836075) support in Skyline to predict retention time of such peptides based on measuring a set of standards on your LCMS set-up. Creating a retention time predictor using SSRCalc in Skyline is described in the Targeted Method Refinement tutorial (pp. 7-10): And also Tutorial 5 of the 2014 SRM Cours at ETH: http://www.srmcourse.ch/downloads.html http://www.srmcourse.ch/tutorials2014/Tutorial-5_Scheduling.pdf Q: How do you target endogenous peptides without a enzyme specificity? Ans: This can be done using a peptide list. Any line-separated set of peptides sequences, regardless of beginning or ending amino acid residues can be pasted directly into Skyline. One such list actually measured with Skyline started like this: RPKPQQFFGLM RPPGFSPFR DRVYIHPFHL ELYENKPRRPYIL ADSGEGDFLAEGGGVR GIGAVLKVLTTGLPALISWIKRKRQQ DRVYIHPF Try pasting this into Skyline, and you will see that it works, just as the list pasted into Skyline during the webinar and then named “Special Peptides” worked. Q: Is there a way to filter the peptide tree for proteotypic peptides queried against a background proteome imported into Skyline? Ans: There is not currently a way to do this in bulk. Originally, we just thought this would be a bad idea, because we saw so many FASTA files with a lot of redundancy, probably stemming from multiple proteoforms of the same gene product. We have learned a lot since then, and Skyline now even queries resources on the web to determine the gene for each protein in a background proteome. We are looking forward to returning to this issue in the near future, and adding new document-wide support for constraining targets to peptides which are unique within the background proteome. For now, Skyline is still limited to the Edit > Unique Peptides form, which acts on a single protein (or peptide list) at a time. Watch the issue below for progress. https://skyline.gs.washington.edu/labkey/issues/home/issues/details.view?issueId=212 Q: Can you use the Background Proteome to warn that a Selected Peptide might have an isobaric partner in the background proteome sharing some of the selected transitions to avoid false positive signals? Ans: Skyline does not support detecting issues of uniqueness at the transition level. We are aware of tools that have done this, but in general think it is not necessarily a good idea. Many hypothetical transitions are not expressed at high enough intensity to cause interference, and to actually cause interference matching transitions must also coelute. If uniqueness were a requirement for chromatography-based quantification, then nothing would be measurable through MS1 filtering, which is obviously not the case. During the time when the CPTAC Verification Working Group was active, we realized that measuring peptides using light and heavy isotope labeled pairs could be susceptible to “auto-interference”, or cases where multiple ions coincided in one form but not the other (e.g. y8++ and b5 having the same m/z in the light form, but not the heavy). We calculated all possible cases of “auto-interferences” and added extra transitions to the measurements in case expected ion abundance ratios were determined to change for these ions (as was expected, if the interference was actually a factor). Note that in this case chromatography is guaranteed to have no impact, because the interference comes from the same peptide. In the end, very few of the hypothetical interferences were determined to actually make an impact. Q: Can you search for Unique peptides in ALL the peptide list and just flag the non-unique ones (instead of looking at one protein at a time?) Ans: This is a great way of asking this question, and it points to an obvious and relatively simple feature that could be implemented in the Edit > Find form in Skyline, where there is an “Advanced” button a list of “custom finders”. Look for a finder to be added to this list soon that finds peptides that are not unique within the background proteome. Once it is added, you will be able to click the “Find All” button to flag these peptides, just as you suggest. Thanks for the great feedback. Q: Do creation of chromatograph library rely on MS data? Ans: Chromatogram libraries are created from prior chromatography-based quantitative experiments. They have the advantage of being able to store relative ion abundance from chromatogram peak areas measured on the same type of instrumentation you will perform future experiments on, and also of being able to store relative peptide precursor ion abundance (with the right experimental setup – see Stergachis, et al. Nature Methods, 2011), and not just fragment ion abundance, as in spectral libraries. A full tutorial on chromatogram libraries can be found on PanoramaWeb: https://panoramaweb.org/labkey/wiki/home/page.view?name=chromatogram_libraries Q: It seems like finding unique peptides is a slow manual process involving deselecting peptides found in other proteins. I am working with parasite targets and want to make sure peptides are not found in common with human proteins or other parasites not of intersest. Would you suggest using the Background proteome for this purpose or should I use another tool outside of skyline? Ans: I would suggest using a background proteome for this. A new feature recently added to the Edit > Unique Peptides form should make this a little easier, but until we improve things further, it will still be necessary to perform this operation on each protein or peptide list separately. If you are measuring many proteins or have many peptide lists, then you are right that it may be more manual than an operation like this should be. Still, I think if you have fewer than 100 proteins, it probably won’t take too long to check them all. Q: Do you know when skyline for metabolomic will be available ? Ans: At least or initial attempts at small molecule support are available now in Skyline-daily, which any Skyline user can request. Just click the green button on the Skyline software web page. The public release will likely happen in a few weeks. Q: Is there a way to just select the 2+ or 3+ precursor depending on which is the highest and to filter this automatically? Thanks! Ans: There are ways to do this in Skyline. One is to measure both and then use the Edit > Refine > Advanced form – Results tab to choose a maximum peptide rank. This will discard all but the peptide precursors with rank less than or equal to the maximum (e.g. for 3, you get 1, 2, 3) This should choose only one precursor per peptide, and only the most intense. I think this is also possible with chromatogram libraries where the peak areas for peptides are specified as being protein relative. Q: I am using .mzxml files from several runs to create a library. The samples I used are different patient samples so many of peptides are overlapped between the runs. What spectra will Skyline use for these overlapping peptides between different .mzxml file and create spectral library? Ans: The current library building algorithm will first choose spectra with the best score in your scoring algorithm. When it finds identical scoring spectra, it will break the tie by choosing the spectrum with the most total signal. We have found this works very well for picking the spectrum at the apex of PRM runs, when we search those PRM runs using spectrum matching software. Q: How does Skyline decide which peptide to pick when there are 2 libraries and both libraries contain that peptide? Thanks! Ans: Skyline will chose the spectrum in based on library order in the Peptide Settings – Libraries tab. It will pick the spectrum from the library that appears highest in that list. Q: Can you manually add transitions when a peptide has not been identified in a spectral library? Ans: Yes. You can manually modify peptides, precursors and transitions by hovering the mouse cursor above the parent element (protein, peptide or precursor respectively) until you see an arrowhead pointing downward to the right of the element text. If you click on this arrowhead, Skyline will show a pop-up pick-list, which you can use to manually alter the selections that Skyline has made automatically based on the Peptide and Transition Settings. These pick-lists were used during the webinar, and they are explained in the Targeted Method Editing tutorial. Q: When you export a report (peptide ratios) is it possible to arrange the report so that that each sample is in a column rather than a row? Ans: Yes. You want to check the Pivot Replicate Name checkbox. This will pivot all per replicate values to the right in separate columns. For instance, if you included just “Peptide Modified Sequence” and “Peptide Retention Time” in your report, and used this option, then you would get a row for each peptide, with the modified sequence in the first column, followed by a column for the retention time in each imported replicate measurement. Q: Is the iRT predictor using RT from the spectral library for the standard peptides versus RT run on the new gradient or how exactly is it predicting the RT? Ans: Hopefully, we will get to do a full webinar on iRT in the future. In the meantime, it is covered in detail in the iRT Retention Time Prediction tutorial. Q: Can you share any acquisition templates for the TTOF 5600? Ans: Unfortunately, we don’t have any. If you post this request to the Skyline support board, we may be able to connect you with people that can help. Q: Is there a way to directly use RT predictions from a Spectrast library? Ans: We have only just been made aware that SpectraST now can include iRT values. At present, Skyline does not use any retention time information from SpectraST libraries, but we expect to support these new features fully in the future. Thanks for asking. |