Ub_dimers_sPRE_XPLOR

Repository containing the code for an upcoming publication.

System requirements

OS: Linux

Software dependencies can be found in env.yml. Rebuild the environment with conda env create -f environment.yml.

Also, encodermap is needed. Install with pip install encodermap.

XPLOR version 3.3

Installation

The paths in the repository have been adjusted to use the repository as a base. To run the code yourself, you need to first clone the repository:

git clone git@github.com:AG-Peter/Ub_dimers_sPRE_XPLOR.git && cd Ub_dimers_sPRE_XPLOR

Next, you need to install the python package containing the helper functions:

pip install -e .

This will add a python-package called xplor to your importable python packages. The install-time of this is less than a minute.

Sample data

A small sample dataset of MD simulations is available under https://sawade.io/Ub_dimers_sPRE_XPLOR/ Please refer to the publication for simulation details and software versions. To get the sample data to test the code yourself, run these commands inside the Ub_dimers_sPRE_XPLOR repository (the download can take a while):

mkdir molsim && cd molsim
wget --recursive --no-parent https://sawade.io/Ub_dimers_sPRE_XPLOR/
mv sawade.io/Ub_dimers_sPRE_XPLOR/K* .
rm -rf sawade.io/

Installation of XPLOR

As this python package heavily relies on the software-package XPLOR, you need to install XPLOR, by following the installation instructions on the XPLOR homepage. The xplor installation comes with a pyXplor executable, that will be used as a secondary python kernel. To make your xplor installation work with the helper functions set the environment variable PYXPLOR_EXECUTABLE before calling any other code (i.e. during your imports):

import numpy as np
import os
os.environ['PYXPLOR_EXECUTABLE'] = '/path/to/pyxplor/executable'
import xplor
xplor.functions.parallel_xplor(['k6'])

The same applies to XPLOR's pdb2psf program. Set it with:

import os
os.environ['PDB2PSF_EXECUTABLE'] = '/path/to/pdb2psf/executable'

Run this Repo

Analysis pipeline

The Analysis pipeline starts with MD simulations. A small subset is available and the full set can be made available upon request. A simulation is defined by a .pdb (topology) and .xtc (trajectory) file. Both of these files should reside in their own directory. The monolithic function xplor.functions.parallel_xplor() takes the argument simdir which should point to the directory containing all the other simulation directories (the determination of this directory is somewhat smart and uses the location from which xplor was installed. eg. /home/user/git/xplor/molsim or /home/user/python_packages/xplor/molsim). It is also recommended to choose a very large subsample (10 000 or even more, otherwise your PC will be busy for quite some time).

df = xplor.functions.parallel_xplor(['k6', 'k29', 'k33'], subsample=10000, parallel=True)

On a machine with a 4-core CPU this command took 5 minutes, 30 seconds to complete. The resulting dataframe can be used to work through the analysis and image creation in the jupyter notebook development/analysis.ipynb.

Starting from MD simulations of the three ubiquitylation sites (K6, K29, and K33).

The analysis of the simulation data was done with the

HDBSCAN

Using HDBSCAN on the full dataset might use a substantial amount of system memory.