Workflow: pxd_to_speclib

Snakemake workflow that reanalyzes proteomics data from a PRIDE Archive project to create a spectral library.

The workflow goes through the following steps:

• Download RAW files from PRIDE Archive for given PXD identifier
• Convert RAW files to MGF using the CompOmics ThermoRawFileParser
• Search with MSGFPlus
• Generate Percolator input files
• Postprocess search results with Percolator
• Parse results to generate a spectral library

Requirements

• Conda (tested on Linux)

Run the workflow

1. Create and activate the environment:

conda env create -f envs/pxd_to_speclib.yml
conda activate pxd_to_speclib

2. Setup your configuration:

   • conf/snakemake_config.json (see Configuration)
   • conf/msgfplus_params.txt
   • Add required input files (e.g. fasta sequence database)

3. Run the workflow:

   • To create a general spectral library: snakemake . --use-conda
   • To create a calibrated retention time dataset: snakemake --snakefile make_rt_lib.smk --use-conda

Configuration

All settings can be found in a JSON configuration file: conf/snakemake_config.json.

Section	Option	Default value	Description
download	pxd_identifier	"PXD000000"	PXD identifier of PRIDE Archive project to download.
	file_pattern	".*"	Regular expression that matches all raw file filenames to download (.* matches all filenames).
convert	exec	"ThermoRawFileParser.sh"	Executable command to call ThermoRawFileParser. See Note 1.
search	msgfplus_conf	"conf/msgfplus_params.txt"	Path to MSGFPlus configuration file.
	fasta	"path/to/search_db.fasta"	Path to protein fasta. Important: MSGFPlus will add decoy peptides by default; they should not yet be present in the given fasta file.
	msgfplus_exec	"msgf_plus"	Executable command to call MSGFPlus. See Note 2.
	threads_per_search	5	Number of threads per MSGFPlus search. See Note 3.

Note 1

ThermoRawFileParser executable
To use a custom version of the ThermoRawFileParser, change convert > exec to:

• Using environment TRFP: "ThermoRawFileParser.sh"
• Using custom TRFP: "mono /path/to/ThermoRawFileParser.exe"

Note 2

MSGFPlus executable
Idem for MSGFPlus, change search > msgfplus_exec to:

• Using environment MSGFPlus: msgf_plus
• Using custom jar file: "msgfplus_exec": "java -Xmx5000M -jar /path/to/MSGFPlus/MSGFPlus.jar"

The latter allows a custom memory limit for the Java VM. By default, this is 1GB.

Note 3

Indirectly limit memory usage while searching with threads_per_search
The option search > threads_per_search defines the number of threads each individual search can use. In combination with the snakemake --cores x option, this allows you to constrict the number of parallel searches. E.g.: the combination of --cores 24 and threads_per_search: 6 limits the number of parallel searches to 4. This can be convenient if you would run into memory issues caused by too many parallel searches.