C- and N-terminal fixed modifications
Viktor
I'm using Crux 1.35 and I'd like to know if there is any
implementation of fixed modifications on C- and N-terminal residues of
peptides in Crux. Fixed modifications exist for individual amino
acids, however, can I do this also for the terminals?
Thank you,
Viktor
B. Frewen
Crux 1.35 does not have fixed modifications on the C- and N-termini, but
it does allow variable modifications of the termini. Use the cmod and
nmod options
http://noble.gs.washington.edu/proj/crux/crux-search-for-matches.html
We plan on adding support for fixed terminal modifications in the next
release. In the mean time, please let me know if you have any more
questions.
Cheers,
Barbara
> --
> You received this message because you are subscribed to the Google Groups "crux-users" group.
> To post to this group, send email to crux-...@googlegroups.com.
> To unsubscribe from this group, send email to crux-users+...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/crux-users?hl=en.
>
>
Viktor Granholm
thank you for your help. However, I have a follow-up question. I'm analyzing iTRAQ-data, which has a 304. 2 Da modification on the N-terminal and on lysine, in my case, on ~ 90 % of the peptides (this is verified by other search engines and other techniques). A variable modification search with Crux finds very few peptides, the ~ 10 % of peptides that doesn't have the modification. Only a very low proportion of these significant PSMs has a modification. Am I using the parameter options (pasted below) incorrectly?
mod=304.199040:K:99
nmod=304.199040:-1
Or could it be some other problem? Thank you again,
Viktor
Hello Viktor,
Crux 1.35 does not have fixed modifications on the C- and N-termini, but it does allow variable modifications of the termini. Use the cmod and nmod options
http://noble.gs.washington.edu/proj/crux/crux-search-for-matches.html
We plan on adding support for fixed terminal modifications in the next release. In the mean time, please let me know if you have any more questions.
Cheers,
Barbara
On Fri, 8 Jul 2011, Viktor wrote:
Dear group,
I'm using Crux 1.35 and I'd like to know if there is any
implementation of fixed modifications on C- and N-terminal residues of
peptides in Crux. Fixed modifications exist for individual amino
acids, however, can I do this also for the terminals?
Thank you,
Viktor
--
You received this message because you are subscribed to the Google Groups "crux-users" group.
To post to this group, send email to crux-...@googlegroups.com.
To unsubscribe from this group, send email to crux-users+unsubscribe@googlegroups.com.
B. Frewen
Looks as though you are using the parameters correctly, so I don't think
that's the problem. You could do a test to help find out what is going
on. Search just a few spectra that you know have modified sequences. You
can use the --scan-number option to search a single spectrum or a range.
Have crux return all of the candidate peptides using the top-match option.
Just set it to some arbitrarily large number. Alternatively, you can
check the matches/spectrum column of search.target.txt to find out how
many candidates those spectra had and set top-match at least that high.
With this test we can see if the correct sequence with the correct
modifications is being considered in the search. If not, there may be a
bug or you may need to modify some other search parameters. If it's there
but not ranking very high, we can dig deeper to find out why.
Let me know if you have questions about this and if I can help interpret
the results.
Thanks,
Barbara
> crux-users+...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/crux-users?hl=en.
>
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "crux-users" group.
> To post to this group, send email to crux-...@googlegroups.com.
> To unsubscribe from this group, send email to
> crux-users+...@googlegroups.com.
Viktor Granholm
thank you for your help a few weeks ago, concerning why Crux reports very few peptides with the fixed iTRAQ modifications, when I use the variable modification parameters.
I followed your advice for troubleshooting, and searched a couple of spectra of which I know the correct peptide sequence (and that they have iTRAQ modifications). Setting the top-match option to 9999, I expect to see all candidate peptides. As you said, with this method we can see if Crux considers the correct peptide, and how it is ranked. For the few example spectra I looked at, I found two problems:
- When the correct peptide was found among the candidates, it was not ranking very high.
- When the correct peptide is fully tryptic, however, it is not found among the candidates when setting the options missed-cleavages=TRUE or digestion=partial-digest. Is this expected? This is likely to explain most of my missed peptides. So, to include both peptides with and without missed cleavages I should run two separate searches?
Thank you.
Viktor
However, the correct peptide were not among the candidate peptides, in the examples I looked at. The correct peptide existed in the database I use, but might not have been found with the digestion rules I used.
--
You received this message because you are subscribed to the Google Groups
"crux-users" group.
To post to this group, send email to crux-...@googlegroups.com.
To unsubscribe from this group, send email to
William Stafford Noble
Hello Barbara,
thank you for your help a few weeks ago, concerning why Crux reports very few peptides with the fixed iTRAQ modifications, when I use the variable modification parameters.
I followed your advice for troubleshooting, and searched a couple of spectra of which I know the correct peptide sequence (and that they have iTRAQ modifications). Setting the top-match option to 9999, I expect to see all candidate peptides. As you said, with this method we can see if Crux considers the correct peptide, and how it is ranked. For the few example spectra I looked at, I found two problems:
- When the correct peptide was found among the candidates, it was not ranking very high.
- When the correct peptide is fully tryptic, however, it is not found among the candidates when setting the options missed-cleavages=TRUE or digestion=partial-digest. Is this expected?
This is definitely not the expected behavior. However, given the first point above, it seems possible that the correct peptide is occurring below 9999 when you expand the search to include missed cleavages or partial digestion. Can you confirm that you are still getting fewer than 9999 candidates in those cases, or can you re-run with --top-match set to something larger?
By the way, how do you know that these identifications are correct?
Thanks.
Bill
2011/7/12 B. Frewen <fre...@u.washington.edu>
--
You received this message because you are subscribed to the Google Groups
"crux-users" group.
To post to this group, send email to crux-...@googlegroups.com.
To unsubscribe from this group, send email to
--
You received this message because you are subscribed to the Google Groups "crux-users" group.
To post to this group, send email to crux-...@googlegroups.com.
To unsubscribe from this group, send email to crux-users+...@googlegroups.com.
Viktor Granholm
the number of matches/spectrum is considerably lower than 99.999: However, I see now that you are still right: despite top-match=99999, only the 501 top PSMs are reported in the output txt-file (I use the sequest-search option). This most likely explains why I couldn't find the correct peptide among the candidates. How should I do to see more than 501 PSMs per spectrum?
About the "correct" identifications:
As this is iTRAQ data, we have run a Mascot search with fixed iTRAQ modifications. Trying to repeat this with Crux, I had to use variable modifications (as there's no fixed N-term modification option). This Crux search found very few peptides, almost all of them without any modification; probably the small proportion of peptides in the experiment that were unlabeled. When I say "correct" peptides I mean a few confident identifications found in Mascot, use to see if they were considered in Crux.
Thank you,
Viktor
B. Frewen
The crux sequest-search command includes a filtering step based on the Sp
score. The default number of matches to keep after filtering is 500,
which is why you are not getting more in your results file. You can set
max-rank-preliminary=99999 to return more candidates with sequest-search
or you can use crux search-for-matches, which does not include the Sp
scoring and filtering.
Barbara
Viktor Granholm
thank you, now I see all peptide candidates, also those that were found to be correct by Mascot. So these peptides are definitely considered by Crux, but they are scored very low.
Viktor
For more options, visit this group
at
http://groups.google.com/group/crux-users?hl=en.
--
You received this message because you are subscribed
to the Google Groups
"crux-users" group.
To post to this group, send email to
crux-...@googlegroups.com.
To unsubscribe from this group, send email to
For more options, visit this group at
http://groups.google.com/group/crux-users?hl=en.
--
You received this message because you are subscribed to the
Google Groups "crux-users" group.
To post to this group, send email to
crux-...@googlegroups.com.
To unsubscribe from this group, send email to
For more options, visit this group at
http://groups.google.com/group/crux-users?hl=en.
--
You received this message because you are subscribed to the Google
Groups "crux-users" group.
To post to this group, send email to crux-...@googlegroups.com.
To unsubscribe from this group, send email to
For more options, visit this group at
http://groups.google.com/group/crux-users?hl=en.
--
You received this message because you are subscribed to the Google Groups
"crux-users" group.
To post to this group, send email to crux-...@googlegroups.com.
To unsubscribe from this group, send email to
For more options, visit this group at
http://groups.google.com/group/crux-users?hl=en.
--
You received this message because you are subscribed to the Google Groups "crux-users" group.
To post to this group, send email to crux-...@googlegroups.com.
To unsubscribe from this group, send email to crux-users+unsubscribe@googlegroups.com.
B. Frewen
I'm glad to hear that at least the candidate peptides are being selected
correctly. I can take a look at your examples to find out why they are
scoring so low. Can you provide me with a few spectra, the fasta file you
are using, and a list of the correct peptide IDs for the spectra? Also,
please let me know what parameters you are using in the search.
Thanks,
Barbara
> � � �� � � � �
> � � ��crux-users+...@googlegroups.com.
> � � �� � � � � �For more options, visit this group
> � � �at
> � � �� � � � �
> � � ��http://groups.google.com/group/crux-users?hl=en.
>
>
>
> --
> You received this message because you are subscribed
> to the Google Groups
> "crux-users" group.
> To post to this group, send email to
> crux-...@googlegroups.com.
> To unsubscribe from this group, send email to
> crux-users+...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/crux-users?hl=en.
>
>
> --
> You received this message because you are subscribed to
> the
> Google Groups "crux-users" group.
> To post to this group, send email to
> crux-...@googlegroups.com.
> To unsubscribe from this group, send email to
> crux-users+...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/crux-users?hl=en.
>
> --
> You received this message because you are subscribed to
> the Google
> Groups "crux-users" group.
> To post to this group, send email to
> crux-...@googlegroups.com.
> To unsubscribe from this group, send email to
> crux-users+...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/crux-users?hl=en.
>
>
> --
> You received this message because you are subscribed to
> the Google Groups
> "crux-users" group.
> To post to this group, send email to
> crux-...@googlegroups.com.
> To unsubscribe from this group, send email to
> crux-users+...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/crux-users?hl=en.
>
>
>
> --
> You received this message because you are subscribed to the Google
> Groups "crux-users" group.
> To post to this group, send email to crux-...@googlegroups.com.
> To unsubscribe from this group, send email to
> crux-users+...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/crux-users?hl=en.
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "crux-users" group.
> To post to this group, send email to crux-...@googlegroups.com.
> To unsubscribe from this group, send email to
> crux-users+...@googlegroups.com.
Viktor Granholm
my problem with ITRAQ searches seems solved now, so I thought I'd sum up the solution for the crux-users group. For ITRAQ searches (which includes modifications on N-terminal and K) the following set of modification parameters works well (cmod is added as K is generally common on the C-terminal):
nmod=304.199040:-1
cmod=304.199040:-1
K=304.199040
Thanks for the help Barbara.
/Viktor
For more options, visit this group
at
http://groups.google.com/group/crux-users?hl=en.
--
You received this message because you are subscribed
to the Google Groups
"crux-users" group.
To post to this group, send email to
crux-...@googlegroups.com.
To unsubscribe from this group, send email to
For more options, visit this group at
http://groups.google.com/group/crux-users?hl=en.
--
You received this message because you are subscribed to
the
Google Groups "crux-users" group.
To post to this group, send email to
crux-...@googlegroups.com.
To unsubscribe from this group, send email to
For more options, visit this group at
http://groups.google.com/group/crux-users?hl=en.
--
You received this message because you are subscribed to
the Google
Groups "crux-users" group.
To post to this group, send email to
crux-...@googlegroups.com.
To unsubscribe from this group, send email to
For more options, visit this group at
http://groups.google.com/group/crux-users?hl=en.
--
You received this message because you are subscribed to
the Google Groups
"crux-users" group.
To post to this group, send email to
crux-...@googlegroups.com.
To unsubscribe from this group, send email to
For more options, visit this group at
http://groups.google.com/group/crux-users?hl=en.
--
You received this message because you are subscribed to the Google
Groups "crux-users" group.
To post to this group, send email to crux-...@googlegroups.com.
To unsubscribe from this group, send email to
For more options, visit this group at
http://groups.google.com/group/crux-users?hl=en.
--
You received this message because you are subscribed to the Google Groups
"crux-users" group.
To post to this group, send email to crux-...@googlegroups.com.
To unsubscribe from this group, send email to
B. Frewen
Just one note on your recommended parameters. The K modification as you
have it is a static mod, not a variable one. That means that ALL K
residues will be modified in the peptides searched. Under some
circumstances, that might not be what you want. A variable
modification of K would be given as
mod=304.199040:K:1
However, you shouldn't need it as the nmod and cmod options should take
care of adding modifications to the ends of peptides. Let me know if you
still have questions.
Thanks for posting this issue.
Barbara
> crux-users+...@googlegroups.com.
> For more options,
> visit this group
> at
>
>
> http://groups.google.com/group/crux-users?hl=en.
>
>
>
> --
> You received this message because
> you are subscribed
> to the Google Groups
> "crux-users" group.
> To post to this group, send email
> to
> crux-...@googlegroups.com.
> To unsubscribe from this group,
> send email to
>
> crux-users+...@googlegroups.com.
> For more options, visit this group
> at
>
> http://groups.google.com/group/crux-users?hl=en.
>
>
> --
> You received this message because
> you are subscribed to
> the
> Google Groups "crux-users" group.
> To post to this group, send email
> to
> crux-...@googlegroups.com.
> To unsubscribe from this group,
> send email to
>
> crux-users+...@googlegroups.com.
> For more options, visit this group
> at
>
> http://groups.google.com/group/crux-users?hl=en.
>
> --
> You received this message because
> you are subscribed to
> the Google
> Groups "crux-users" group.
> To post to this group, send email
> to
> crux-...@googlegroups.com.
> To unsubscribe from this group,
> send email to
>
> crux-users+...@googlegroups.com.
> For more options, visit this group
> at
>
> http://groups.google.com/group/crux-users?hl=en.
>
>
> --
> You received this message because
> you are subscribed to
> the Google Groups
> "crux-users" group.
> To post to this group, send email
> to
> crux-...@googlegroups.com.
> To unsubscribe from this group,
> send email to
>
> crux-users+...@googlegroups.com.
> For more options, visit this group
> at
>
> http://groups.google.com/group/crux-users?hl=en.
>
>
>
> --
> You received this message because you
> are subscribed to the Google
> Groups "crux-users" group.
> To post to this group, send email to
> crux-...@googlegroups.com.
> To unsubscribe from this group, send
> email to
> crux-users+...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/crux-users?hl=en.
>
>
> --
> You received this message because you
> are subscribed to the Google Groups
> "crux-users" group.
> To post to this group, send email to
> crux-...@googlegroups.com.
> To unsubscribe from this group, send
> email to
> crux-users+...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/crux-users?hl=en.
>
>
> --
> You received this message because you are subscribed to the Google
> Groups "crux-users" group.
> To post to this group, send email to crux-...@googlegroups.com.
> To unsubscribe from this group, send email to
> crux-users+...@googlegroups.com.
Lukas
Just as you say we would like to treat all the iTRAQ labels as fix
modifications. We would hence be very interested in seing a posibility
to process data using a fix N-terminal modification. The current
nmod=304.199040:-1 only allows for a variable mod, which inflates the
search space.
Thanks
-Lukas
> > <thaba...@gmail.com>
> ...
>
> read more »
William Noble
Crux release. Note that Barbara is out of till mid-next week, so we
won't have movement on this immediately.
Bill
>> read more �
Viktor
We found a related question when Barbara looked through this ITRAQ
issue. She found that adding a variable modification on the C-terminal
was necessary for the ITRAQ search, despite that ITRAQ-mods are only
found on N-term and K. It seemed as if variable and fixed
modifications on K weren't considered for C-terminal K, so a variable
C-term mod had to be added. It would be good to make sure all amino
acids modifications are also applied when found on the terminals.
Thank you,
Viktor
>
> read more »