Re: [spctools-discuss] Re: No prot-MODELS
Hi Luis, I suspect that my problem may be due to multiple installations of TPP (from this post): https://groups.google.com/forum/#!msg/spctools-discuss/Tq-i-7SVoJE/4XZcq4EGCgAJ;context-place=forum/spctools-discuss I had originally installed 5.1 in a preferred location, but parts of TPP didn't run - so I uninstalled and installed at default path. Unknowingly I've done all my analysis in the location of what was the old installation. The temporary fix in the above thread didn't work for me when I tried to repeat it (substituting in the appropriate path). Please could you tell me is there a fix that you'd recommend? or will have I reinstall? (I'd prefer not to if possible - there's several days worth of searches etc in there and I'm concerned I'll mess up file paths). - would it help if I sent over an example interact.prot.xml file? Thanks Pete On Tuesday, 12 June 2018 15:47:47 UTC-7, pbell@gmail.com wrote: > > Hi Luis, > > I've got the same problem with my dataset (also a large dataset, combining > multiple experiments). - No models html file is generated. > > Please can you tell me if this issue was resolved? > > Thanks > Pete > > > On Monday, 7 May 2018 19:01:28 UTC-7, Luis wrote: >> >> Hi Heeyoun, >> >> Does it seem that the file is complete when you open it in the >> ProtXMLViewer? (can you see the last protein group, for example?) If so, >> maybe there is an issue with the script that generates the models html >> page. Are you able to share the interact.prot.xml file so I can test it >> locally? If so, feel free to email it to me directly. >> >> Cheers, >> --Luis >> >> >> On Mon, May 7, 2018 at 5:44 PM, Heeyoun Hwang wrote: >> >>> Hi, Luis >>> >>> Thank you for your kindness. >>> >>> I opened the ProtXMLViewer with interact.prot.xml file, but >>> interact.prot-MODEL.html file was not automatically generated. >>> >>> The MODEL.html file is not exist, so I cannot checked that is corrupt or >>> broken and I cannot delete it. >>> >>> In other words, PepXMLViewer can generate interact.pep-MODEL.html file. >>> PepXMLViewer is fine. >>> >>> This interact.prot.xml has 47k protein entries. Is this too many number >>> to generate the model? >>> >>> Best, >>> >>> Heeyoun >>> >>> >>> >>> >>> 2018년 5월 8일 화요일 오전 7시 24분 29초 UTC+9, Luis 님의 말: Hello, The MODELS.html file should be auto-generated when you open the ProtXMLViewer (if one does not exist). If for some reason the models file already exists but is somehow not displaying or is corrupt/broken, the easiest thing to do is to delete that models file and re-open the ProtXMLViewer. Let us know if that still fails, and of any errors that you may notice in this case. Hope this helps, --Luis On Fri, May 4, 2018 at 2:55 AM, Heeyoun Hwang wrote: > Hi, all again, > > Is there anybody to help me, please? > > I want to get prot-MODELS.html file which was not automatically > generated. > > Thank you. > > Heeyoun Hwang > > 2018년 5월 1일 화요일 오후 5시 12분 52초 UTC+9, Heeyoun Hwang 님의 말: > >> Hi, >> >> I have to use TPP for large DATA sets including 24 RAW files. >> >> Most search process have been worked well, but prot-MODELS.html file >> is not generated. >> >> Surely, interact.prot.xml and its index is good. >> >> How can I get the MODELS.html file? >> >> Regards, >> >> Heeyoun Hwang >> >> >> >> -- > 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 post to this group, send email to spctools...@googlegroups.com. > Visit this group at https://groups.google.com/group/spctools-discuss. > For more options, visit https://groups.google.com/d/optout. > -- >>> 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 post to this group, send email to spctools...@googlegroups.com. >>> Visit this group at https://groups.google.com/group/spctools-discuss. >>> For more options, visit https://groups.google.com/d/optout. >>> >> >> -- 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 post to this group, send email to spctools-discuss@googlegroups.com. Visit this group at https://groups.google.com/group/spctools-discuss. For more options, visit https://groups.google.com/d/optout.
Re: [spctools-discuss] Valid combination of multiple samples, fractions and digestions with different enzymes, using prophets
Please note - I've discovered a similar problem (no models html file generated) in the following thread: https://groups.google.com/forum/#!searchin/spctools-discuss/interact$20prot%7Csort:date/spctools-discuss/Yrg2yEO4otg/8u0bNKzpAAAJ Cheers Pete On Monday, 11 June 2018 14:58:53 UTC-7, pbell@gmail.com wrote: > > Hi David, > > Thanks for your advice - I've had success using the experiment flag in > Petunia, and have found that using an experiment label including fraction, > replicate and also enzyme (used for digest) works well when processed > through iProphet/protein prophet. > > However, my problem now is that the 'models' tab when I open the resulting > interact.ipro.prot.xml file is not visible - so I don't know where to set > the cutoff for 1% FDR > > I notice in my .params folder that there is no interact.ipro.prot html > file - but no errors appear while protein prophet is running. > > Please could you shed any light on how I can make the models visible? > Perhaps I've deleted some file in the .params folder that's required to > generate the html? > > Thanks > Pete > > > > On Wednesday, 6 June 2018 00:34:47 UTC-7, David Shteynberg wrote: >> >> Hello Pete, >> >> iProphet has a sibling experiments model and uses the replicate spectra >> model for replicate PSMs that are in the same experiment.This is enabled by >> running InteractParser with -X flag, which labels the >> spectra in the pepXML file. If you are using xinteract or Petunia web >> interface the option is -E. You have to make sure that >> for each search engine analysis you assign the same label to the same >> data. The experiment label is flexible and allows you to separate the data >> into "experiments" as defined by you. It makes sense in your case to make >> the experiment labels either the "fraction_name" or the >> "fraction_name"+"replicate". Other than that I think you are on the right >> path. >> >> Cheers, >> -David >> >> On Tue, Jun 5, 2018 at 3:13 PM, wrote: >> >>> Hi, >>> >>> I'd really appreciate advice regarding the most valid way to combine my >>> searches with peptide / i / protein prophet. >>> >>> I have 3 samples, 3 fractions per sample, and each fraction was digested >>> with multiple enzymes. Each of these digests were injected twice. >>> >>> The resulting data were then searched with different search engines; all >>> in an attempt to increase number of protein IDs. >>> >>> My idea of the workflow was as follows: >>> >>>1. combine results of 1 search engine for duplicate injections of a >>>single fraction using peptide prophet >>>2. combine results of multiple search engines using iprophet >>>3. combine iprophet results from different enzymatic digestions of a >>>single fraction of a single sample using protein prophet (to group >>> sibling >>>peptides) >>> >>> I'm unclear whether/when it is valid for me to combine: >>> a) different fractions (note - fractions are expected to have some >>> overlap in peptide and protein IDs) >>> b) different samples (note- samples are biological replicates, and are >>> expected to contain the same peptides / proteins) >>> >>> The reason I would like to combine them all together, is so that I can >>> have a single protein FDR for the whole experiment. >>> >>> Thanks! >>> Pete >>> >>> >>> >>> -- >>> 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 post to this group, send email to spctools...@googlegroups.com. >>> Visit this group at https://groups.google.com/group/spctools-discuss. >>> For more options, visit https://groups.google.com/d/optout. >>> >> >> -- 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 post to this group, send email to spctools-discuss@googlegroups.com. Visit this group at https://groups.google.com/group/spctools-discuss. For more options, visit https://groups.google.com/d/optout.
Re: [spctools-discuss] Re: No prot-MODELS
Hi Luis, I've got the same problem with my dataset (also a large dataset, combining multiple experiments). - No models html file is generated. Please can you tell me if this issue was resolved? Thanks Pete On Monday, 7 May 2018 19:01:28 UTC-7, Luis wrote: > > Hi Heeyoun, > > Does it seem that the file is complete when you open it in the > ProtXMLViewer? (can you see the last protein group, for example?) If so, > maybe there is an issue with the script that generates the models html > page. Are you able to share the interact.prot.xml file so I can test it > locally? If so, feel free to email it to me directly. > > Cheers, > --Luis > > > On Mon, May 7, 2018 at 5:44 PM, Heeyoun Hwang > wrote: > >> Hi, Luis >> >> Thank you for your kindness. >> >> I opened the ProtXMLViewer with interact.prot.xml file, but >> interact.prot-MODEL.html file was not automatically generated. >> >> The MODEL.html file is not exist, so I cannot checked that is corrupt or >> broken and I cannot delete it. >> >> In other words, PepXMLViewer can generate interact.pep-MODEL.html file. >> PepXMLViewer is fine. >> >> This interact.prot.xml has 47k protein entries. Is this too many number >> to generate the model? >> >> Best, >> >> Heeyoun >> >> >> >> >> 2018년 5월 8일 화요일 오전 7시 24분 29초 UTC+9, Luis 님의 말: >>> >>> Hello, >>> >>> The MODELS.html file should be auto-generated when you open the >>> ProtXMLViewer (if one does not exist). If for some reason the models file >>> already exists but is somehow not displaying or is corrupt/broken, the >>> easiest thing to do is to delete that models file and re-open the >>> ProtXMLViewer. >>> >>> Let us know if that still fails, and of any errors that you may notice >>> in this case. >>> >>> Hope this helps, >>> --Luis >>> >>> >>> On Fri, May 4, 2018 at 2:55 AM, Heeyoun Hwang >>> wrote: >>> Hi, all again, Is there anybody to help me, please? I want to get prot-MODELS.html file which was not automatically generated. Thank you. Heeyoun Hwang 2018년 5월 1일 화요일 오후 5시 12분 52초 UTC+9, Heeyoun Hwang 님의 말: > Hi, > > I have to use TPP for large DATA sets including 24 RAW files. > > Most search process have been worked well, but prot-MODELS.html file > is not generated. > > Surely, interact.prot.xml and its index is good. > > How can I get the MODELS.html file? > > Regards, > > Heeyoun Hwang > > > > -- 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 post to this group, send email to spctools...@googlegroups.com. Visit this group at https://groups.google.com/group/spctools-discuss. For more options, visit https://groups.google.com/d/optout. >>> >>> -- >> 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 post to this group, send email to spctools...@googlegroups.com >> . >> Visit this group at https://groups.google.com/group/spctools-discuss. >> For more options, visit https://groups.google.com/d/optout. >> > > -- 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 post to this group, send email to spctools-discuss@googlegroups.com. Visit this group at https://groups.google.com/group/spctools-discuss. For more options, visit https://groups.google.com/d/optout.
Re: [spctools-discuss] iProphet for Leucine/isoleucine IDs
Yes your understanding of the issue is spot on. iProphet converts everything to L for comparison, then reports the top hit from the highest scoring search engine as reported by the search engine. Post processing has to take care of how to treat the variants. Cheers, -David On Tue, Jun 12, 2018 at 11:53 AM, Chih-Chiang Tsou < chihchiang.t...@gmail.com> wrote: > Hi David, > > Thank you for the quick reply. The three search engines I am using they > rank I/L variant peptides inconsistently. Some ranks Leucine-version as > rank 1, some ranks Isoleucine-version as rank 1. > > Then in iProphet pepXML file I observed that for each spectrum iProphet > picks rank 1 hit info from either one of the three search results (the one > with largest PeptideProphet probability?) to report in iProphet PepXML. > Therefore for some spectrum, Leucine-version is reported, but sometimes > Isoleucine-version gets reported. > > I was concerned about whether the inconsistent ranking would affect > iProphet scoring, but then as you mentioned, iProphet converts I/L variant > and treat them as the same sequence therefore it should be fine as far as > scoring. Just in the end when we read iProphet result we need to also treat > I/L variants properly. Am I understanding it correctly? > > Best, > Chih-Chiang > > > On Tue, Jun 12, 2018 at 12:26 PM David Shteynberg systemsbiology.org> wrote: > >> Hello Chih-Chiang, >> >> iProphet will convert all Isoleucines reported to Leucines for comparison >> purposes. Have you observed an error in your analyses? >> >> -David >> >> On Tue, Jun 12, 2018 at 9:06 AM, Chih-Chiang Tsou < >> chihchiang.t...@gmail.com> wrote: >> >>> Dear TPP developers, >>> >>> I have a question regarding how iProphet handles ambiguous I/L IDs from >>> different search engine. I am using X!Tandem, Comet, and MSGF+ and combine >>> the results using iProphet, and noticed that the three search engines have >>> different rules ranking I/L hits. >>> So how does iProphet handle such case if different search engines give >>> different I/L variant as top ranking PSM? Does iProphet threat different >>> I/L variant peptide as different IDs and therefore penalizing the final >>> probability score, or iProphet considers them as the same ID? >>> >>> Thanks, >>> Chih-Chiang >>> >>> >>> >>> >>> -- >>> 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 post to this group, send email to spctools-discuss@googlegroups.com. >>> Visit this group at https://groups.google.com/group/spctools-discuss. >>> For more options, visit https://groups.google.com/d/optout. >>> >> >> -- >> You received this message because you are subscribed to a topic in the >> Google Groups "spctools-discuss" group. >> To unsubscribe from this topic, visit https://groups.google.com/d/ >> topic/spctools-discuss/VDEKdnL5iEQ/unsubscribe. >> To unsubscribe from this group and all its topics, send an email to >> spctools-discuss+unsubscr...@googlegroups.com. >> To post to this group, send email to spctools-discuss@googlegroups.com. >> Visit this group at https://groups.google.com/group/spctools-discuss. >> For more options, visit https://groups.google.com/d/optout. >> > -- > 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 post to this group, send email to spctools-discuss@googlegroups.com. > Visit this group at https://groups.google.com/group/spctools-discuss. > For more options, visit https://groups.google.com/d/optout. > -- 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 post to this group, send email to spctools-discuss@googlegroups.com. Visit this group at https://groups.google.com/group/spctools-discuss. For more options, visit https://groups.google.com/d/optout.
Re: [spctools-discuss] iProphet for Leucine/isoleucine IDs
Hi David, Thank you for the quick reply. The three search engines I am using they rank I/L variant peptides inconsistently. Some ranks Leucine-version as rank 1, some ranks Isoleucine-version as rank 1. Then in iProphet pepXML file I observed that for each spectrum iProphet picks rank 1 hit info from either one of the three search results (the one with largest PeptideProphet probability?) to report in iProphet PepXML. Therefore for some spectrum, Leucine-version is reported, but sometimes Isoleucine-version gets reported. I was concerned about whether the inconsistent ranking would affect iProphet scoring, but then as you mentioned, iProphet converts I/L variant and treat them as the same sequence therefore it should be fine as far as scoring. Just in the end when we read iProphet result we need to also treat I/L variants properly. Am I understanding it correctly? Best, Chih-Chiang On Tue, Jun 12, 2018 at 12:26 PM David Shteynberg < david.shteynb...@systemsbiology.org> wrote: > Hello Chih-Chiang, > > iProphet will convert all Isoleucines reported to Leucines for comparison > purposes. Have you observed an error in your analyses? > > -David > > On Tue, Jun 12, 2018 at 9:06 AM, Chih-Chiang Tsou < > chihchiang.t...@gmail.com> wrote: > >> Dear TPP developers, >> >> I have a question regarding how iProphet handles ambiguous I/L IDs from >> different search engine. I am using X!Tandem, Comet, and MSGF+ and combine >> the results using iProphet, and noticed that the three search engines have >> different rules ranking I/L hits. >> So how does iProphet handle such case if different search engines give >> different I/L variant as top ranking PSM? Does iProphet threat different >> I/L variant peptide as different IDs and therefore penalizing the final >> probability score, or iProphet considers them as the same ID? >> >> Thanks, >> Chih-Chiang >> >> >> >> >> -- >> 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 post to this group, send email to spctools-discuss@googlegroups.com. >> Visit this group at https://groups.google.com/group/spctools-discuss. >> For more options, visit https://groups.google.com/d/optout. >> > > -- > You received this message because you are subscribed to a topic in the > Google Groups "spctools-discuss" group. > To unsubscribe from this topic, visit > https://groups.google.com/d/topic/spctools-discuss/VDEKdnL5iEQ/unsubscribe > . > To unsubscribe from this group and all its topics, send an email to > spctools-discuss+unsubscr...@googlegroups.com. > To post to this group, send email to spctools-discuss@googlegroups.com. > Visit this group at https://groups.google.com/group/spctools-discuss. > For more options, visit https://groups.google.com/d/optout. > -- 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 post to this group, send email to spctools-discuss@googlegroups.com. Visit this group at https://groups.google.com/group/spctools-discuss. For more options, visit https://groups.google.com/d/optout.
Re: [spctools-discuss] iProphet for Leucine/isoleucine IDs
Hello Chih-Chiang, iProphet will convert all Isoleucines reported to Leucines for comparison purposes. Have you observed an error in your analyses? -David On Tue, Jun 12, 2018 at 9:06 AM, Chih-Chiang Tsou wrote: > Dear TPP developers, > > I have a question regarding how iProphet handles ambiguous I/L IDs from > different search engine. I am using X!Tandem, Comet, and MSGF+ and combine > the results using iProphet, and noticed that the three search engines have > different rules ranking I/L hits. > So how does iProphet handle such case if different search engines give > different I/L variant as top ranking PSM? Does iProphet threat different > I/L variant peptide as different IDs and therefore penalizing the final > probability score, or iProphet considers them as the same ID? > > Thanks, > Chih-Chiang > > > > > -- > 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 post to this group, send email to spctools-discuss@googlegroups.com. > Visit this group at https://groups.google.com/group/spctools-discuss. > For more options, visit https://groups.google.com/d/optout. > -- 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 post to this group, send email to spctools-discuss@googlegroups.com. Visit this group at https://groups.google.com/group/spctools-discuss. For more options, visit https://groups.google.com/d/optout.
[spctools-discuss] iProphet for Leucine/isoleucine IDs
Dear TPP developers, I have a question regarding how iProphet handles ambiguous I/L IDs from different search engine. I am using X!Tandem, Comet, and MSGF+ and combine the results using iProphet, and noticed that the three search engines have different rules ranking I/L hits. So how does iProphet handle such case if different search engines give different I/L variant as top ranking PSM? Does iProphet threat different I/L variant peptide as different IDs and therefore penalizing the final probability score, or iProphet considers them as the same ID? Thanks, Chih-Chiang -- 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 post to this group, send email to spctools-discuss@googlegroups.com. Visit this group at https://groups.google.com/group/spctools-discuss. For more options, visit https://groups.google.com/d/optout.
