Using spectral-counts

After you have run a search and assigned confidence to your identified PSMs, you can now calculate relative quantification values for the proteins by running the command:

$ crux spectral-counts --threshold 0.1 --protein-database small-yeast.fasta ./crux-output/

While the command is running, you will see output like this

INFO: Beginning spectral-counts.
INFO: Total proteins found: 56
INFO: Assigning index 0 to ./crux-output/
INFO: Number of matches:109
INFO: Number of matches passed the threshold 61
INFO: Number of peptides 15
INFO: Number of proteins 13

When the command is finished, it will print

INFO: Elapsed time: 0.0122 s
INFO: Finished crux spectral-counts.

Within the crux-output directory, there will be three new files:

  • – tab-delimited file containing the spectral-counts results
  • spectral-counts.params.txt – the parameters used for running spectral-counts
  • spectral-counts.log.txt – the log output of spectral-counts
  • The file will look something like:

    protein id 	NSAF
    YLR043C 	0.61643416
    YGL135W 	0.17810018
    YEL027W 	0.10352074
    YGL009C 	0.035437137
    YLR185W 	0.031369921
    YJR069C 	0.014012959
    YGR192C 	0.0083149187
    YMR235C 	0.0067826854
    YBR118W 	0.0060274079

    In this output, the first field is the identifier and the second field is the spectral-counting measure (NSAF, dNSAF, SIN, or EMPAI) result for each protein. To select which measure is computed, see the crux spectral-counts documentation.

    Using peptideprophet probability from pepXML

    crux spectral-counts also supports using Peptide Prophet probabilities provided in a pepXML file. Use the command:

    $ crux spectral-counts --protein-database small-yeast.fasta --threshold-type custom --custom-threshold-name peptideprophet --threshold 0.9 --custom-threshold-min F target.pep.xml
    This command will filter out matches whose Peptide Prophet probability is < 0.9 and calculate NSAF for the proteins of the remaining PSMs.

    Using mzIdentML files with spectral-counts

    crux spectral-counts also supports mzIdentML as input. For example,

    $ crux spectral-counts --measure RAW file.mzid

    If the protein sequences are not provided within the given mzIdentML file, then you will have to provide a database using the protein-database parameter for the NSAF, dNSAF, EMPAI, and SIN metrics.