ResidueFisher

ResidueFisher is a bioinformatics protocol for the search of protein homology using a three-step ”search, detect, and enrich” model that uses a combination of structural and sequence aligners working in tandem to filter and enrich conservation signals, resulting in "fished" residues that may shed light into understanding ans studying a protein of interest.

Table of Contents

• Dependencies
• Installation
• Usage
• Support
• Contributing

Dependencies

ResidueFisher depends on:

• bash
• conda
• docker
• mafft
• tmux

Installation

To install ResidueFisher, run the following code snippet:

git clone https://github.com/cvigilv/ResidueFisher
cd ResidueFisher
make configure
conda activate ResidueFisher

In order to use ResidueFisher, the conda environment must be active (conda activate ResidueFisher)

Usage

Database preparation

To prepare a database using Foldseek, run the script bin/prep_database.sh script as follows:

sh bin/prep_database.sh <FOLDSEEK-DATABASE-NAME> <INTERNAL-DATABASE-NAME>

# Example preparation of PDB dataset available in Foldseek
sh bin/prep_database.sh PDB mypdb

To see the available datasets, run bin/prep_database.sh without arguments.

To prepare a database from PDB files, please refer to foldseek tutorial. In order for ResidueFisher to work correctly, user created databases must be inside a directory named data/foldseek_dbs in the proyect root and must contain FASTA files for the aminoacid sequence and 3di sequence, which can be generated as follows:

foldseek convert2fasta <USER-DB-NAME> <USER-DB-NAME>.fasta
foldseek lndb <USER-DB-NAME>_h <USER-DB-NAME>_ss_h
foldseek convert2fasta <USER-DB-NAME>_ss <USER-DB-NAME>_ss.fasta

Query protein preparation

Unlike Foldseek, this protocol is intended to study a single protein chain; therefore, in order to use ResidueFisher, one must first extract a chain of interest from its original PDB file.

In the src/scripts folder, there is a script called splitchains.py, which extracts all the chains from a particular PDB file and saves them as separate files for use in ResidueFisher.

Note: the recommended way of preparing and storing all the structure files is creating a new folder in data called queries (due to the nature of this files) and run the chain splitting script inside this folder. Here is an example of this procedure:

# Ensure we have the conda environment activated 
conda activate ResidueFisher

# Assuming you are at the project root...
mkdir data/queries
cd data/queries
wget https://files.rcsb.org/download/3F3P.pdb
python ../../src/scripts/splitchains.py 3F3P.pdb

From this example, a total of 13 should be found inside the data/queries folder: 1 for the original structure and 12 corresponding to the chains A through L of 3F3P.

Foldseek-fishing Usage

To use ResidueFisher, run the script bin/ResidueFisher script as follows:

sh bin/ResidueFisher <PDB-FILE> <INTERNAL-DATABASE-NAME>

# Example using the previously prepared protein and dataset
sh bin/ResidueFisher data/queries/3F3P_C.pdb mypdb

This will generate a folder in results with the following structure:

results/3F3PC_pdb/
├── foldseek/
├── msa/
├── moma/
└── tree/

Inside each subdirectory, log files and result can be found in order to analyse and study the protein used in the protocol.

Support

Please open an issue for support.

Contributing

Please contribute using Github Flow. Create a branch, add commits, and open a pull request.

License

MIT