HDXer

Introduction

HDXer is a Python package that can be used to:

• Compute Hydrogen-Deuterium exchange (HDX) data from an atomistic ensemble of protein structures (e.g. from molecular dynamics simulations)
• Reweight a candidate ensemble by applying a maximum-entropy biasing scheme, so that the computed HDX data conform to a target experimental set of HDX-MS data, within a defined level of error

---

Requirements

• Python 3 (preferably Anaconda Python 3.x version)
• git
• NumPy
• MDTraj
• matplotlib
• pytest
• JupyterLab
• HDXer

---

Installation

You should run all the code shown below on terminal using Bash. Bash is a command language interpreter that's already available on Linux/Mac operating systems. See the Bash tutorial if you are not familiar with Bash.

If you are using a Windows computer, you will have to download and use Git for Windows instead of a terminal. Git for Windows will let you use both Bash and git on a Windows computer.

Python 3

For Python 3, we recommend Anaconda Python 3.x version, a free and open-source distribution that comes with useful standard Python libraries. You will be able to download, access, and manage Python packages more effectively by using conda, a package manager within Anaconda. Ananconda Python 3.x version can be downloaded from the Anaconda website.

A user guide for Anaconda and Conda cheat sheet are great resources if you are new to using Anaconda and want to learn more about it.

git

git is a free and open-source distributed version-control system.

Again, if you are using a Windows computer, download Git for Windows. Some operating systems (e.g. MacOS or Linux) may have git installed already. You can check this by running the following command:

git --version

If there is no git available on your computer, you can install it using conda, or your operating system's package manager.

conda install -c conda-forge git

If you encounter errors at this point, or using the conda commands below, you may need to update your $PATH environment variable and/or initialize conda on your system. The Anaconda package will usually provide an initialization script for this purpose, for example, on Windows available in:

~/Anaconda/etc/profile.d/conda.sh

You can call this script or other initialization command in your ~/.bashrc file (for example, by running echo ". ~/Anaconda/etc/profile.d/conda.sh" >> ~/.bashrc) to make sure it's run every time you open a new Bash terminal. Check out the Anaconda documentation for your filesystem if you continue to have problems, as conda commands are required for the remaining parts of the installation.

HDXer Python Package

Download the HDXer Python package using git clone.

git clone https://github.com/Lucy-Forrest-Lab/HDXer

Once the HDXer Python package is downloaded, create a new conda environment that includes all of the dependencies.

cd HDXer
conda env create -f HDXER_ENV.yml

Every time you use HDXer, you have to activate the HDXER_ENV using conda activate.

conda activate HDXER_ENV

Within the HDXER_ENV environment, you have all the dependencies available to install the HDXer Python package to your Python environment. HDXer is still in development, so we recommend installing the package in 'editable' mode, using either pip:

cd ..
pip install -e HDXer

or conda:

cd ..
conda develop HDXer

Installing in 'editable' mode will allow you to pull updates directly from this Github repository to your local HDXer installation, without having to reinstall the package.

Now, you finished installing the HDXer Python package. Let's add the path to the HDXer directory as in your .bashrc and/or .bash_profile. The HDXer directory will be used throughout the tutorials and referred to as $HDXER_PATH.

cd HDXer

echo -e "\nexport HDXER_PATH='${PWD}'" >| ~/.HDXER_PATH_variable

grep -qxF "if [ -f ~/.HDXER_PATH_variable ]; then source ~/.HDXER_PATH_variable; fi" ~/.bashrc || echo -e "if [ -f ~/.HDXER_PATH_variable ]; then source ~/.HDXER_PATH_variable; fi" >> ~/.bashrc
source ~/.bashrc

grep -qxF "if [ -f ~/.HDXER_PATH_variable ]; then source ~/.HDXER_PATH_variable; fi" ~/.bash_profile || echo -e "if [ -f ~/.HDXER_PATH_variable ]; then source ~/.HDXER_PATH_variable; fi" >> ~/.bash_profile
source ~/.bash_profile

You can easily access the HDXer directory using the $HDXER_PATH variable. For example, to move to the HDXer directory, you simply have to type the following command on terminal:

cd $HDXER_PATH

---

Testing

After you've carried out the steps above, we recommend that you run the functional tests provided with HDXer to check for a successful installation. You can also use this as an opportunity to check your Bash and Conda environments have been successfully created.

First, check the folder $HDXER_PATH/HDXer/tests/data/tmp_test_output is empty. This is where the test outputs will be stored, so remove any files from previous test runs before running the tests anew. Then, to run the tests, open a new Bash shell and type:

cd $HDXER_PATH
conda activate HDXER_ENV
pytest -v

The tests will take roughly 60 seconds to complete. All tests should pass. If you encounter any failures, you can check the output of the tests in the folder where they're stored ($HDXER_PATH/HDXer/tests/data/tmp_test_output) to help determine the cause of the error. If the failure persists and you believe there's a problem with HDXer, please contact the developers or raise an issue on this GitHub repository.

---

Tutorials

The tutorials for the HDXer are available in a series of easy-to-follow Jupyter notebooks. These notebooks can be viewed within the GitHub page or with Jupyter Lab. You will be able to run code interactively within each notebook using Jupyter Lab. Run the following commands on terminal to access the notebooks using Jupyter Lab:

cd $HDXER_PATH/tutorials/notebooks
jupyter lab

In the notebooks/ directory, there are five Jupyter notebooks.

• 01_data_prep.ipynb
• 02_calc_hdx.ipynb
• 03_reweighting.ipynb
• 04_decision_plot.ipynb
• 05_heatmap.ipynb

These notebooks will walk you through how to run HDXer and how to analyze the results with an example application of both HDX calculations and ensemble reweighting.

N.B. As per the Jupyter website, the Jupyter Notebook aims to support the latest versions of these browsers:

• Chrome
• Safari
• Firefox

Up to date versions of Opera and Edge may also work, but if they don’t, please use one of the supported browsers.

---

Copyright

Copyright (c) 2020, 2021, Richard T. Bradshaw, Fabrizio Marinelli, Paul Suhwan Lee

Acknowledgements

Project based on the Computational Molecular Science Python Cookiecutter version 1.1.