TTT Proteotyping Pipeline¶

Contents:

Installation¶

TTT Proteotyping Pipeline consists of several steps:

raw to mzXML conversion with ReAdW.

peptide identification with X!Tandem.

X!Tandem xml output to FASTA conversion with convert_tandem_xml_2_fasta.py.

Peptide to reference database alignment with BLAT.

Antibiotic resistance detection with Proteotyping.

Taxonomic composition estimation with Proteotyping.

Listing distinct/unique proteins in X!Tandem output with create_unique_protein_list.py.

ReAdW, X!Tandem, and BLAT are external programs. This documentation will give brief instructions on how to install and use them together with TTT Proteotyping Pipeline. The Python programs are a part of this Python package, available in the TTT_proteotyping_pipeline folder in the repository.

The steps can be run manually, or preferably via the included Snakemake script.

Download the code¶

To download the code, clone the repository:

$ hg clone https://bitbucket.org/chalmersmathbioinformatics/TTT_proteotyping_pipeline

This will clone the entire repository to a folder called TTT_proteotyping_pipeline in your current directory.

ReAdW¶

ReAdW is meant to be run under Windows, but can be run under Linux using Wine, see instructions below. Running ReAdW in Wine requires a Linux system with a working 32-bit Wine installation.

Note

It is important that you use 32-bit Wine, as ReAdW cannot be run under 64-bit Wine. As of this writing, 32-bit Wine is only available for RedHat Enterprise Linux 6 and below. Support for 32-bit Wine was removed in RHEL 7.

Get ReAdW¶

ReAdW can be downloaded from the ReAdW Github repository. Either clone the entire repository or download the binary suitable for your system. Note the information about the dependencies on three Windows DLL files: XRawfile2.dll, fileio.dll, fregistry.dll. These files are NOT supplied with this pipeline.

Create a 32-bit Wine prefix¶

Install Wine. It is important that a 32-bit version of Wine is installed, this normally means packages named <package>.i686 instead of <package>.x86_64. In RHEL/CentOS it can be installed like this:
```
yum install wine
```
Create a Win32 prefix from which to run ReAdW. Make sure to set and export WINEARCH=win32 during the creation of the wine prefix. Modify the command below to a path of your choice. Note that this step likely requires working X11-forwarding:
```
export WINEARCH=win32
export WINEPREFIX=/path/to/your/desired/wineprefix
winecfg
```
Click OK in any configuration windows that pop up.
Download winetricks to install the required Visual Studio C++ runtimes. vcrun2010 is required for ReAdW and vcrun2008 is required for the Thermo DLL’s. Again, note that this step requires X11-forwarding to be enabled:
```
wget https://raw.githubusercontent.com/Winetricks/winetricks/master/src/winetricks
sh winetricks vcrun2010 vcrun2008
```

Click through any installation prompts that pop up, and after they complete, finish by registering XRawfile2.dll in your wine prefix:
wine regsvr32 XRawfile2.dll

Running ReAdW¶

Make sure to set the WINEPREFIX environment variable to the correct path (same directory you specified when creating the 32 bit wine prefix), then run ReAdW from your Linux command prompt via Wine:

export WINEPREFIX=/path/to/your/desired/wineprefix
wine /path/to/ReAdW.201510.xcalibur.exe [options] /path/to/sample.raw

Now it should run.

X!Tandem¶

Download and install X!Tandem according to the instructions on the X!Tandem homepage. To run a sample in X!Tandem, several xml-files must be prepared. There is a Python program in TTT Proteotyping Pipeline called run_xtandem.py that will automatically create the required input files and run X!Tandem for you. It makes it very easy to use X!Tandem:

run_xtandem.py --output OUTFILE --db /PATH/TO/FASTA --threads N --xtandem /PATH/TO/TANDEM.EXE

Running¶

The TTT Proteotyping Pipeline can be run either by manually running each of the steps, or it can be controlled automatically using Snakemake. The automation ensures that each RAW input file is taken through all the required steps to produce the final output. Together with the supplied snakemake_crontab_script.sh it can be used as a completely hands-off automated way of analyzing proteomics samples.

Note

The Snakemake workflow can only be run on Linux computers, as it depends on some Linux command line features.

Work directory¶

The Snakemake workflow requires a work directory containing the following folder structure:

raw
mzXML
xml
fasta
blast8
results

The reference data required to run the entire workflow is usually put in a single directory (or symlinked there), but they can (in theory) be located anywhere in the file system. The position of all the required files must be specified in the TTT_pipeline_snakemake_config.yaml file. This file must be specified on the command line when invoking the workflow.

Run the Snakemake workflow¶

To run the Snakemake workflow, ensure that a suitable Python/Conda environment is activated in which all the proteotyping programs and scripts are available in PATH. The minimal command line required to start the workflow is this:

snakemake --snakefile SNAKEFILE --configfile CONFIGFILE

As the work directory is specified in the configfile, the command can in theory be run anywhere in the file system. It is recommended, however, that the Snakemake workflow is invoked via the use of the included snakemkae_crontab_script.sh which sets some environment parameters to ensure reliable operation. It uses the linux command flock to ensure that only one instance of the workflow is ever run at the same time.

Automatic invokation via crontab¶

The workflow can be invoked automatically at set times using the Linux built-in crontab. To edit your personal user’s crontab, type crontab -e at the command prompt. Add something like the following line to make the Snakemake workflow check for new files to analyze three times daily (00:00, 12:00, 18:00):

0 0,12,18 * * * /bin/bash /PATH/TO/snakemake_crontab_script.sh

Make sure to modify the configfile (TTT_pipeline_snakemake_config.yaml) and the crontab script file (snakemake_crontab_script.sh) to match your environment.