VaxDesign

Script that computationally designs a vaccine.

Requirements

1. Download this script:

wget https://raw.githubusercontent.com/sarisabban/VaxDesign/master/VaxDesign.py

2. Use the following commands (in Ubuntu GNU/Linux) to install all nessesary programs and Python libraries for this script to run successfully:

sudo apt update && sudo apt install gnuplot dssp python3-bs4 python3-biopython python3-lxml -y

3. Make sure you install PyRosetta as the website describes.

Description

This script autonomously designs a vaccine from a user specified target site. It is best to use structures that contain a target site bound to a receptor or an antibody since these structures will be used to guide the vaccine development. Once you identify a target site, this script will graft it , without the need for you to intervene, onto a scaffold protein structure and design it resulting in an ideal protein structure displaying your target site in its original 3D cofiguration. Thus the protien, theoretically, can be used as a vaccine against this site, and hopefully neutralise the disease you are researching. Running the script willr esult in multiple protein structures with different sequences. It is best to use these structures and evaluate them (such as forward fold them using the abinitio protocol) to choose the best one to synthesise.

This script has been last tested to work well with PyRosetta 4 Release 223 using Python 3.7. If you use this script on a newer PyRosetta or Python version and it fails please notify me of the error and I will update it.

The script does the following:

Scaffold Search

1. Chooses ideal scaffold structures from a scaffold database.

Vaccine Design

1. Isolates the motif.

2. Isolates the receptor.

3. Grafts the motif onto a specified scaffold.

4. Performs the Fold From Loop protocol to optimise the grafted structure (NOT IMPLEMENTED YET).

5. RosettaDesign the structure around the motif to find the sequence that would fold the proteins to its designed structure.

6. Generate fragments for Rosetta AbinitioRelax folding simulation and analyses their quality.

Output files are as follows for each generated structure:

	File Name	Description
1	frags.200.3mers	3-mer fragment of sequence designed structure from the Robetta server
2	frags.200.9mers	9-mer fragment of sequence designed structure from the Robetta server
3	grafted.pdb	Grafted motif to scaffold structure
4	motif.pdb	Original requested motif
5	pre.psipred.ss2	PSIPRED secondary structure prediction of sequence designed structure from the Robetta server
6	plot_frag.pdf	Plot of the fragment quality RMSD vs Position
7	receptor.pdb	Original receptor that binds motif
8	scaffold.pdb	Scaffold structure
9	structure.fasta	Fasta of Rosetta Designed structure
10	structure.pdb	Final RosettaDesigned vaccine structure

Manual

Autonmous Structure Generation

1. Go to Robetta server and register a username, it is very easy and straightforward.

2. Search potential protein scaffolds using the following command:

python3 VaxDesign.py -s PDBID RCHAIN CHAIN FROM TO DATABASE use -s or --scaffold

• PDBID = The protein's Protein Data Bank identification name
• RCHAIN = The chain where the receptor resides within the protein .pdb file
• CHAIN = The chain where the target site resides (not part of the receptor) within the protein .pdb file
• FROM = The start of the target site
• TO = The end of the target site
• DATABASE = The directory that contains multiple .pdb scaffold structures to search through

3. The script will generate a directory called Scaffolds which will contain all the scaffold structures. Search through the directory and choose one scaffold (move it to the working directory next to the VaxDesign.py script).

4. Generate a vaccine using the following command:

python3 VaxDesign.py -p PDBID RCHAIN CHAIN FROM TO SCAFFOLD PROTOCOL USERNAME use -p or --protocol

• PDBID = The protein's Protein Data Bank identification name
• RCHAIN = The chain where the receptor resides within the protein .pdb file
• CHAIN = The chain where the target site resides (not part of the receptor) within the protein .pdb file
• FROM = The start of the target site
• TO = The end of the target site
• SCAFFOLD = The protein scaffold structure generated from the Scaffold.py script
• PROTOCOL = Choose between fixbb ot flxbb RosettaDesign
• USERNAME = The Robetta server username to generate and download the AbinitioRelax fragment files and the PSIPRED secondary structure prediction file

Calculation time is about 12 hours on a normal desktop computer. Access to the internet is a requirement since the VaxDesign.py script will be sending and retrieving data from some servers. The script will result in 20 vaccine structures.

5. Use this Rosetta Abinitio script to simulate the folding of all final designed vaccine structures. An HPC (High Preformance Computer) and the original C++ Rosetta are required for this step. This is just an evaluation step to check the design before you synthesise it, other folding simulation algorithms can be used, use what you find most reliable.

Manual Structure Generation

You can run each step separatly, use the following commands to run each step:

• Isolate motif:

python3 VaxDesign.py -m PDBID RCHAIN FROM TO use -m or --motif

• Isolate receptor:

python3 VaxDesign.py -r PDBID RCHAIN use -r or --receptor

• Graft motif onto the scaffold structure

python3 VaxDesign.py -g RECEPTOR.pdb MOTIF.pdb SCAFFOLD.pdb use -g or --graft

• Run the Fold From Loop protocol (Not Implemented Yet)

python3 VaxDesign.py -f motif.pdb grafted.pdb FROM TO USERNAME use -f or --ffd

In this case the FROM and TO are the positions of the motif on the grafted structure and not the original protien structure

• RosettaDesign the structure

python3 VaxDesign.py -d PROTOCOL grafted.pdb FROM TO use -d or --design

In this case the FROM and TO are the positions of the motif on the grafted structure and not the original protien structure. PROTOCOL can be either fixbb or flxbb, with fixbb being the reccomended one.

You can also choose to RosettaDesign only the suface of the structure (without changing your motif), use surface as the protocol.

• Generate fragments

python3 VaxDesign.py -F structure.pdb USERNAME use -F or --fragments

• Help menu

python3 VaxDesign.py -h use -h or --help

Tutorial

Here is a tutorial that walks you through how to use the script and the results that can be expected. Here is also a video that performs this tutorial

Search a database of structures for a scaffold: python3 VaxDesign.py -s 2y7q A B 332 337 Scaffold_Database

Run the entire protocol: python3 VaxDesign.py -p 2y7q A B 332 337 scaffold.pdb fixbb siwa2 where scaffold.pdb is a cleaned structure of the PDB ID 3HZ7

Breakdown of the protocol:

Isolate the motif: python3 VaxDesign.py -m 2y7q B 332 337

Isolate the receptor: python3 VaxDesign.py -r 2y7q A

Graft the motif onto the scaffold using the receptor as reference: python3 VaxDesign.py -g receptor.pdb motif.pdb scaffold.pdb

Fold From Loop protocol (Not Implemented Yet): python3 VaxDesign.py -f motif.pdb grafted.pdb 61 66 siwa2

Sequence design the grafted structure: python3 VaxDesign.py -d fixbb grafted.pdb 61 66

Generate fragments for the grafted structure (for an Abinitio simulation): python3 VaxDesign.py -F structure.pdb siwa2

Sequence design of the surface only: python3 VaxDesign.py -d surface grafted.pdb 1 3 5 7 8 9 10 11 12 13 16 17 19 20 23 24 25 27 28 37 38 39 40 43 47 50 51 52 53 54 55 57 58 59 60 73 74

Reference

Please reference the following when using this script.

• Sari Sabban (2020) Computationally grafting an IgE epitope onto a scaffold. bioRxiv 2020.09.14.296392; doi: https://doi.org/10.1101/2020.09.14.296392
• Sari Sabban (2021) Computationally grafting an IgE epitope onto a scaffold: Implications for a pan anti-allergy vaccine design. 19(ISSN 2001-0370) pp. 4738-4750