SA2py ( Solvent-Accessible-Surface-Area in Python)

Program to calculate the Solvent-Accessible Surface Area (SASA) of a protein.
Give the the area in Square Angstrom for each atom and for the protein.
This program is devide in three scripts python :

• main.py, to run the program
• sasa.py, with all functions to compute the SASA
• sasa_stat.py, with all functions to build statisticals graphs or pymol files

The methode use to compute the SASA is based about this article :
Environment and Exposure to Solvent of Protein Atoms. Lysozyme and Insulin by A.Shrake and J A Rupley

Setup your environnement

Clone the repository

git clone https://github.com/ElouanBethuel/SASA.git

Move to the project directory

cd SASA

Install conda
Install mamba

conda install mamba -n base -c conda-forge

Create and activate the conda environnement

conda env create -f sasa_project.yml
conda activate sasa_project 

Run the program

Run with the program :

python main.py 7kh5 100

Execute the script main.py with the pdf file 7kh5.pdb and with 100 points to model the solvation sphere.
You can replace 7kh5 by any another PDB ID. This program work with many PDB files.If the PDB file format is not compatible an error message is raise.

To create also graphics and pymol files for visualisation add the argument -s

python main.py 7kh5 100 -s 

Outputs

IIn the shell, the Solvent Accessible Surface Area of the protein is displayed :

Calculates the solvent accessible surface area from the 7kh5 PDB file :

Downloading the PDB file	
Loading the PDB file

Downloading PDB structure '7kh5'...


Calculation ongoing, please wait a few seconds...

=================
Final result :	
The solvent accessible surface area of the protein : 6503.51 Ų 

Files generated

With by default arguments :

• The pdb file
• A text file with the solvent accessibility of each atom

With additional arguments (-s) :

• A png file (graph) to visualize the points created around a atom
• A png file (graph) showing solvent accessibility by atom category
• A png file (graph) showing solvent accessibility by amino acid category
• A pymol file for showing the accessibility of the protein surface
• A pymol file for showing the neighbors selection

This files are stored in two folders

• Outputs : for stored all outputs
• PDB : for stored all pdb files

The file name format is always the same (exemple with the 7kh5 pdb ID)

• pdb file : pdb7kh5.ent (pdb + pdb_id + .ent)
• png file : pdb_id + name_file + .png
• texte file : pdb_id + .txt
• pymol file (neihgbors) : pdb_id + _neighbors.pse
• pymol file (surface) : pdb_id + _surface.pse

To visualize pymol files with the shell

pymol outputs/7kh5_surface.pse 

This image shows the 7kh5 protein with its surface coloured according to its accessibility to the solvent: the blue parts are the least accessible and the red ones the most accessible.

Results

Comparison of our program results with those of PDBePISA

According to the website https://www.ebi.ac.uk/pdbe/pisa/ : PDBePISA (Proteins, Interfaces, Structures and Assemblies) is an interactive tool for the exploration of macromolecular interfaces.

With 20 points by atoms to model the solvation sphere :

PDB ID	SASA	PDBePPISA	percentage difference
3i40	3644.85 Ų	4902.6 Ų	34.5 %
7kh5	6821.96 Ų	6346.1 Ų	6.9 %
3wgt	28621.66 Ų	31243.3 Ų	8.1 %
3rt9	21911.93 Ų	22296.2 Ų	1.7 %
1us7	21904.68 Ų	23019.5 Ų	4.8 %

With 400 points by atoms to model the solvation sphere :

PDB ID	SASA	PDBePISA	Percentage difference
3i40	3727.46 Ų	4902.6 Ų	31.5 %
7kh5	6563.24 Ų	6346.1 Ų	3.3 %
3wgt	29599.49 Ų	31243.3 Ų	5.5 %
3rt9	22202.34 Ų	22296.2 Ų	0.42 %
1us7	22182.53 Ų	23019.5 Ų	3.7 %

We can see that the higher the number of points, the better the performance.
In addition, performance is much better for medium and large proteins than for small ones (3i40 for example).

Comparison of our program results with the scientific paper

Comparaison of the solvent accessible surface area per residus with 7kh5 :

With SASA program :

With data of the article :

In the article, the solid line corresponds to solvent accessible surface area per residue, and we can see that the results are similar to those of our program. For example, In both cases, we can see a important accessibility between residues 10 and 15 and a import decrease of the accessibility between residues 50 and 61.

---