Methods and Objectives

Python version support ============== Python 2.7, 3.4 and higher

Required

• Biopython
• matplotlib
• NumPy
• pandas
• pytz
• PyYAML
• requests

Recommended

• coverage

Installation with Anaconda ============== - Begin by cloning the git repository. - If desired, create a virtual environment. (This is the easiest and cleanest way to satisfy dependencies.) - Update the PATH variable on your system. - Finally, run driver_parse.py to begin data download and analysis.

Create Virtual Environment -----------------Create a virtual environment with Anaconda using the conda package manager:

conda create --name pdb_34 python=3.4.3 biopython requests pyyaml pytz pandas coverage

If using an IDE, remember to update your settings to use the new virtual environment. Otherwise, on Windows activate the environment with the command:

conda activate pdb_34

On linux, activate the environment with the command:

source activate pdb_34

Update PATH Variable -----------------Update your system path variable with the location of the PDB and lib directories as shown below.

Windows

• Non-persistent PATH update (disappears on reboot):

  SET PATH="%path%;C:\\YourInstallationPath\\PDB";C:\\YourPath\\PDB\\pdb\\lib"

• Persistent PATH update (Stackoverflow): http://stackoverflow.com/a/28545224/3182836

Linux/OSX (POSIX)

• Non-persistent PATH update (disappears on reboot):

  export PATH=$PATH:/YourInstallationPath/PDB:/YourInstallationPath/PDB/pdb/lib

• Persistent PATH update: Add the export PATH command to ~/.profile

This script generates 2 tab delimited files that have the composite of secondary structure, missing regions, and missing regions assignment as outlined in the paper Resolving the Ambiguity: Making Sense of Intrinsic Disorder when PDB structures disagree (in review). The first file has PDB composite information, and it will be created in the folder working/ with the filename pdb_seq.tsv. The second file has the UniProt composite information and it is stored in the uni_data folder/ and it is called uni_composite.<date identifier>.json/yaml/tsv. In order to use these for analysis, these files can be loaded into Pandas dataframes, or another analysis tool. The yaml or json format preserves the data structures within.

driver_parse.py is the analysis orchestrator and the entry point into the application. These scripts will retrieve any current data files needed and create working data folders in the home directory. driver_parse does not currently accept any command-line arguments, therefore, to begin analysis run driver_parse as follows:

python driver_parse.py

Logging options may be configured in logging_config.yaml.

The following logs are written to the user home directory:

• pdb.log: General application message.
• uni_download_errors.log: Errors encountered when downloading FASTA files.
• missing_uniprots.log: Any missing UniProt IDs due to download errors.

The application will use existing data files when present. Therefore, on subsequent runs, remove any data files that you would like to have re-downloaded.

Note that if errors are encountered, or processing is interrupted, driver_parse.py may be re-run and will resume at the furthest possible point. If errors are encountered during one of these follow-up runs, the applicaiton may be trying to use zero-byte data left over from the previous run. Therefore remove that file before continuing. Once removed, the application will re-download the file and resume processing. This will be handled automatically in a future release.