Basically I need a way to enter a sequences and find out if either all of it or peptides derived from it are present in the Human proteome (either MS-MS data or anything else).
For example: I have the sequence ASDFGHKLCVNMWERTYIP. i want to find if 'NMWERTYIP' is found, or 'ASDFGHKLCV' or maybe 'LCVNMWERTY' or even 'FGHKLCVNMW' all are 100% identical but are only partial porions of the sequence I entered.
Any available database (searchable)? please don't say BLAST because i am looking for NOVEL sequences.
Thanks
Dear Rajendra Mandage,
What I need is exactly the opposite. I have predicted peptides that do not correlate to any known or predicted (even hypothetical) proteins in any database (NCBI BLAST nr and others). I am trying to compare their sequences and or parts of their sequences to mass-spec spectras that might indicate the potential of these novel sequences to actually being transcribed and translated as proteins, even though they were never predicted or discovered by any conventional methods...
Is there anything like this available (example: the seuqence ADSFGHLKLLKMNVCVVCMNPIIPIYTRTETWQRT: I am looking if any of part of this peptide was actually found in any MS-MS analysis although it is not related to any known or predicted protein:
If "ADSFGHLKLL" was found that would be great, but not if it was found as a part of ADSFGHLKLLGGRTY... (because that would indicate that the full sequence I put above is not really a true novel sequence...
If found some serial peptides like:
ADSFGHLKLL
HLKLLKMNVCVV
ADSFGHLKLLKMN
CMNPIIPIYTRTETWQ
CMNPIIPIYTRTETWQR
CMNPIIPIYTRTETWQRT
That is even better...
Dear Ronen,
We currently have a method for this in house at the CompOmics group. We basically perform an in depth analysis on all spectra present in the PRIDE repository. We do a sequence database search against all of these experiments using the standard UniProt human proteins + their isoforms. We also take into account common contaminants (mainly to ensure your novel peptides are not matching by chance and could have a better match against a known item). For example, what you consider a spectrum matching a novel peptide might in fact have a better match against pig trypsin...
If you are interested, feel free to contact me !
Cheers,
Kenneth
Also look at our new neXtProt tools, the Unicity Checker:
https://www.nextprot.org/tools/unicity-checker
To our knowledge its the only tool that will also take into account about 2 million sequence variants so that if your peptide exist in a modified form in one of the reported SNP or in a cancer mutation (COSMIC) you will also find it.
- Badges
- Science topic
- Similar topics
- Biological Science
- Biochemistry
- Biomolecules
- Peptides