Protein Structure Proximity Network

This project aims to create a network out of a protein structure stored in the standard PDB format.

The proximity network (in this context) of a protein is defined as a graph $G=(V, E)$ where $V$ is the set of amino acid residues (indexed) within the protein sequence and $E$ is the set of edges connecting these residues. An edge is established between two residues when they are close to each other, i.e., euclidean distance is small and will be taken $<8Å$ by default.

Installation

First clone this repository locally & enter it:

bash git clone https://github.com/raysas/protein-structure-proximity-network.git cd protein-structure-proximity-network

Dependencies:    

Install the following dependencies through pip:

• Biopython
• NetworkX
• Pyvis
• Pandas
• Numpy
• Seaborn
• Matplotlib

bash pip install -r requirements.txt

Running the code

1. Through terminal (original)

The workflow takes from a use a pdb id to generate the network. All that has to be done is to run the following command on the terminal:

bash python code/generate_network.py <pdb_id>

Or if you wish to set the distance threshold to a different value, you can run:

bash python code/generate_network.py <pdb_id> <distance_threshold>

Make sure you input a valid PDB id. Check here for more info. You can also provide a pdb file path instead of an id, under testing - succesful trials so far.

2. Import python module

If you wish you can run each function from extracting the pdb file to visualzing tee network by importing the python module and calling the functions.

```python import sys sys.path.append('/code')

from generate_network import * ```

3. Run the Jupyter notebook on Colab

You can also run the code on a Jupyter notebook (this file) on Google Colab. Either add this repository to you google drive or connect it to github (might give issues without saving the PDB file anywhere).

**p.s.* this is a work in progress and the code is not yet optimized. More features to be added soon listed in the to-do list*

Output

In the data folder you will find a new directory of named by pdb_id after running

data
└── pdbid
        ├── output.log
        ├──pdbid.pdb
        ├──pdbidcontactmap.png
        ├──pdbidtnetwork.graphml
        └──pdbidtnetworkviz.html

The output.log file contains the log of the process. The pdbid.pdb file is the PDB structure of the protein. The pdbidcontactmap.png is the contact map of the protein. The pdbidtnetwork.graphml is the network in graphml format. The pdbidtnetwork_viz.html is the interactive visualization of the network generated from pyvixz.

Example run: building a network for 6xdc

bash python code/generate_network.py 6xdc

The following changes will be made in the data folder:

After retrieving the PDB structure, it will extract residues and coordinates information to generate a contact map. The contact map is a heatmap of the distance between residues. Note that the coordinates for each residues are defined by the coordinates of their alpha carbon.

The network is then generated (by default threshold distance=8) and saved in graphml format. The network is also visualized in an interactive html file. The layout is set based on the x and y coordinates for each residues (plane z=0).

Extensions

Subsequent analysis and visualization using softwares and tools like networkx, PyG, Gephi and CytoScape can be done out of the generated .graphml network file.

Recent works have been using protein structural networks to apply deep learning teachniques like Graph Neural Networks (GNNs) and Graph attention Networks (GATs) (nature paper reference).

References

Hamelryck, T., Manderick, B. (2003) PDB parser and structure class implemented in Python. Bioinformatics 19: 2308–2310

Hagberg, A., Swart, P. J., & Schult, D. A. (2008). Exploring network structure, dynamics, and function using NetworkX (No. LA-UR-08-05495; LA-UR-08-5495). Los Alamos National Laboratory (LANL), Los Alamos, NM (United States).

License

This repository is licensed under the MIT License, contributions are welcome! <!-- wanna mention citation file --> Don't forget to cite this repository if you use it in your work:

bibtex @software{Adam_Protein_structure_proximity, author = {Adam, Rayane}, license = {MIT}, title = {{Protein structure proximity network generator}}, url = {https://github.com/raysas/protein-structure-proximity-network}, version = {1.0.0} }