Comet release 2017.01

    release 2017.01 rev. 0 (2017.01.0), release date 2017/??/??
  • Major update to code to add PSI's enhanced fasta format (PEFF) support. Comet currently supports the ModRes and VariantSimple keywords only. This enables one to search annotated variable modifications and amino acid substitutions. The ModRes modifications can be analyzed in conjunction with the standard Comet variable modifications.
    • Note that the controls for Comet's standard variable modifications (such as binary mods, force modification, etc.) only apply to the standard variable modifications and not the PEFF mods.
    • For the VariantSimple amino acid substitutions, Comet will currently only allow a single amino acid substitution in a peptide at a time.
    • PSI-Mod OBO entries must have "xref: DiffMono:". UniMod OBO entries must have "xref: delta_mono_mass:". Entries without these values are ignored.
    • Addition of source="peff" or source="param" attributes to the mod_aminoacid_mass element in the pepXML output.
    • Addition of id attribute, referencing the modification ID from the OBO file, to the mod_aminoacid_mass element in the pepXML output.
    • PEFF modifications will not be specified in the "search_summary" element at the head of each pepXML file.
    • Note that any static mods are *always* applied and variable or PEFF modification masses are added on top of static mods.
    • When a PEFF modification occurs at the same position as a PEFF variant, the modification will be considered at that position on both the original residue as well the variant residue.
    • The .out, .sqt, and .pin output formats currently expect a modification character (e.g. * or #) to appear in the sequence for each variable modification. This does not extend well to PEFF modifications so for these output formats, the peptide string will replace these modification characters with the bracketed modification mass difference e.g. "[79.9663]". This will break many downstream tools that expect modification characters but unfortunately there's no other solution.
  • In pepXML modification output, 'static="mass"' and 'variable="mass' are added to each "mod_aminoacid_mass" element.
  • Added a "verbose_output" hidden parameter. This parameter controls whether or not (default) extra diagnostic information is reported. Currently it warns if a PEFF modification does not have a mass difference entry or if a PEFF variant is the same residue as in the original sequence. This parameter is "hidden" in that it is not written to the comet.params file generated by "comet -p" nor is it present in the example comet.params files available for download. To invoke this parameter, a user must manually add "verbose_output = 1" to their comet.params file.
  • Comet now tracks and reports all duplicate protein entries. The residues I and L treated as being the same/equivalent. You can turn off I/L equivalence by adding an undocumented/hidden parameter "equal_I_and_L = 0".
  • Allow the specification that a variable modification cannot occur on the C-terminal residue. This is accomplished by setting the 5th parameter entry in the variable_mods" parameter to "-2".
  • Limit the number of permutations of modifications within a single peptide to 10,000. This limit is separately applied to both variable modifications and PEFF modifications.
  • Change Crux compiled text output to also represent mass differences (e.g. M[15.9949]) instead of actual modified masses (e.g. M[147.0]) in the modified peptide string. This brings it in line with the standard text output. Additionally, both standard and Crux versions will now report the mass differences to one decimal point.
  • Modified the options for the "isotope_error" parameter. Please see the parameters page for the changes.
  • Fixes recently discovered bug. In releases 2015.01.0 through 2016.01.2, any peptide with a variable modification had its precursor mass calculated with the mass types assigned to the fragment ions. So if one specified average masses for the precursor and monoisotopic masses for the fragments, all modified peptides would have incorrect peptide masses (calculated with monoisotopic mass values). This bug would not be relevant if both mass types were assigned to the same type i.e. monoisotopic masses used for both precursor and fragment ions. Thanks to D. Zhao for identifying the bug.
  • Change how the "scan_range" parameter is applied. One can set either the start scan or end scan independently now. For example, "scan_range = 5000 0" will search all scans starting at 5000. Similarly "scan_range = 0 250" will search all scans from the first scan to scan 250. Previously, a start scan of 0 would ignore this parameter and a scan range setting of "5000 0" would not search as the end scan is less than the start scan. This also fixes the issue where no spectra would be searched if the specified first scan was not the appropriate scan type (i.e. MS/MS scan for "scan_level = 2"). Reported by A. Sharma.
  • I'm considering phasing out support for writing ".out" files. The parameter entry "output_outfiles" is no longer documented online nor written in the example params files. However, it is still functional if you add "output_outfiles = 1" manually into your comet.params files. If you still use ".out" files and would like support for that format continued, let me know.

Documentation for parameters for release 2017.01 can be found here.

Go download from the download page.