WEBnma3 API

This is the source code for calculating normal modes and performing further analyses used on WEBnma web server.

• WEBnma v3 API is a refactoring work of WEBnma v2 API, which means:

  • v2 and v3 provide the same functions, but are implemented in different code structures. In particular, v2 uses Python2, and v3 uses Python3 with part of the code being parallelized to make use of modern multicore CPUs.
  • the core function, normal mode calculation, in v2 is provided by MMTK, while in v3 it is re-implemented with Python3.

• When using WEBnma v2 or v3, please cite:

  Tiwari SP, Fuglebakk E, Hollup SM, Skjærven L, Cragnolini T, Grindhaug SH, Tekle KM, Reuter N. WEBnm@ v2.0: Web server and services for comparing protein flexibility. BMC Bioinformatics. 2014; 15:427

Install:

For MacOS, Linux:

[Optional] If you are not using mamba instead of conda, we recommend running the following for much faster conflict-resolution and environment setup in conda.

  conda install -n base conda-libmamba-solver
  conda config --set solver libmamba

There are two simple setup alternatives. The first is to install with conda, using

conda install -c bioconda -c salilab -c ddx webnma

Alternatively, it is also easy to install from source:

1. Download the repository, e.g. using git clone https://github.com/reuter-group/webnma3.git.

2. Navigate to the new directory that is created, and enter conda env create --file environment.yml.

   NB! If you are using an Apple M1 processor (for which some library dependencies are not yet available), you should probably run instead: CONDA_SUBDIR=osx-64 conda env create --file environment.yml.

3. Still on that terminal window, activate the new environment and setup webnma:

   conda activate webnma3-env
   python setup.py install

Done! See the usage instructions below.

For Windows users:

WEBnma does not have a conda build for Windows unfortunately, because some a key dependency (MUSTANG) do not have one. One way to get around this might be to use WSL.

You are welcome to contribute your solution here if any.

Usage:

1. Use webnma via the web interface: WEBnma web server
2. Or use webnma on your own machine after installing webnma successfully. Type webnma from your command line and you will see the usage instruction:

   
   $ webnma
   usage: webnma [-h] {mode,eigen,fluc,mode_vis,corr,sa,dl,ca} ...

WEBnma v3 normal mode calculation and analyses

positional arguments: {mode,eigen,fluc,mode_vis,corr,sa,dl,ca} mode Calculate normal modes eigen Calculate average deformation energies and plot eigenvalues fluc Calculate fluctuations and atomic displacement analysis mode_vis Calculate protein trajectories for specified mode numbers/range corr Perform correlation analysis sa Perform normal mode analyses for a single structure dl Download protein structure files from PDBe; support chain selection ca Perform comparative analyses for multiple structures

optional arguments: -h, --help show this help message and exit

To see how to use each sub-command/analysis you can input `webnma <sub-command> -h`, e.g `webnma mode -h` to see how to run a normal mode calculation.

More usage examples:

$ webnma dl 1su4 # download the file pdb1su4.ent from PDBe

$ webnma mode pdb1su4.ent # calculate the (default 200) normal modes for this structure

$ webnma sa 1su4 # perform all the default analyses for the single structure 1su4

# Dependencies:
- Python packages:
  - numpy
  - scipy
  - numba
  - biopython
  - matplotlib
  - seaborn
- [MUSTANG](https://lcb.infotech.monash.edu/mustang/)

PDB files used in WEBnma 3 are downloaded from PDBe(Protein Data Bank in Europe).