ligQplus

Massive compound screening for pathogenic proteomes.

Work scheme

In order to carry out this pipeline, the first thing to do is create a conda environment where you can install a series of necessary packages. See here more information about Anaconda. In addition, it is necessary to have a version of python 3 at least.

Download and pre-requisites

$ git clone https://github.com/BIAlabIC/ligQplus.git

Create execution environment:

$ conda create --name ligQ_plus

Activate environment

$ conda activate ligQ_plus

Once in the ligQ plus environment install:

$ conda install -c bioconda blast
$ pip install urllib3==1.25.8
$ pip install tqdm
$ pip install pandas
$ pip install Bio
$ pip install chembl_webresource_client
$ pip install numpy
$ pip install bioservices
$ pip install sqlalchemy 
$ pip install rdkit

In order for your Linux system to find the given command/executable file, we need to make sure to run the following command in the using terminal before starting work:

$ export PYTHONPATH=$PYTHONPATH:/path/to/ligQplus/
$ export PATH=$PATH:/path/to/ligQplus/

Create databases

In addition to the scripts, it is necessary to have a series of Databases. Some scripts search using databases locally via SQL such as ChEMBL and others via URL/API. Some databases should be previously generated to save computing time. It is necessary to have the space to store these databases.

MOAD

The every_bind.csv file is generated based on experimental assays and contains curated information about interactions between ligands and proteins available in the PDB database.

$ wget -O moad.csv "http://bindingmoad.org/files/csv/every_bind.csv"

$ python3 /patho_pdb_domain/MOAD_PDBIND/MOAD.py -i moad.csv > moad.json

CHEMBL

The ChEMBL database can be at least 20gb in size, please check if you have enough space before downloading.

Note: The stable version for this pipeline is ChEMBL 28. If you want to use it another version of this database, please let us know us if you have any issues.

$ wget ftp://ftp.ebi.ac.uk/pub/databases/chembl/ChEMBLdb/#last_version#/

From this base, you must create 2 tables that summarize the information of PFAM domains per target molecule and the ligands that each presents:

$ python3  /path/to/ligqplus/patho_chembl/pfam_trg_sql_assay.py -db chembl_#.db > pfam_assay_##.csv

$ python3  /path/to/ligqplus/patho_chembl/pfam_trg_sql_mech.py -db chembl_#.db > pfam_mech_#.csv

Mapping pfam-pdb and pdb-ligands

The pdb_pfam_mapping.txt file contains information about the PFAM domains (start/end positions) of each of the protein structures deposited in the PDB.

The lig_pairs.lst file contains information about the ligands co-crystallized with each of the protein structures deposited in the PDB.

$ wget -O pdb_pfam_mapping.txt "http://ftp.ebi.ac.uk/pub/databases/Pfam/mappings/pdb_pfam_mapping.txt"

$ wget -O lig_pairs.lst "http://www.ebi.ac.uk/thornton-srv/databases/pdbsum/data/lig_pairs.lst"

Prepare proteome and extract Uniprot IDs

Note: "Proteome_uniprot.fasta" refers to all proteins in uniprot DataBase related to the organism of interest. Must be download from Uniprot. If you like you can use our test file to check.

$ makeblastdb -in Proteome_uniprot.fasta -dbtype prot

$ blastp -db Proteome_uniprot.fasta -query organism.fasta -qcov_hsp_perc 80 -num_threads 4 -max_hsps 1 -outfmt 6 > file.txt

$ tr '.' ',' < file.txt | awk '$3>=90' > organism_filter.txt (filter by identity)

$ awk -F " " '{print $1, "", $2}' organism_filter.txt > organism_final.txt (keep the first 2 columns)

Note: If the previously executed command doesn't work (losing identity values of 100.000), perform the filtering without changing "." to ",".

'organism_final.txt' It will finally have 2 columns. The first containing the locus tag of each protein of the organism of interest and the second column with the respective UniprotID. There can be multiple Uniprot IDs for each locus tag.

Scripts:

note: Make sure to have all the files, DataBases and scripts in the same folder before proceeding. Including: organism/XXX_final.txt, pfam_assay_##.csv, pfam_mech_##.csv, chembl_##.db, moad.json, etc.

Run

1. prepocessing_proteome_liqQ.py

  $ python3 prepocessing_proteome_liqQ.py -i /path/to/organism/organism_final.txt -o organism -n organism

2. Create organism directory (mkdir organism inside path/to/organism/)

• make_dic.py (create directories with the list of Uniprot IDs to test)

  $ python3 make_dic.py -i organism_lt.txt (output from 1.) -o organism/organism/

3. proteome_compound.py (returns a list of ligands by Uniprot ID)

  $ python3 proteome_compound.py -i organism_lt.txt (output from 1.) -o organism/organism/

4. analysis_result.py (Takes the previous output and combines the results by ChEMBL and PDB)

  $ python3 analysis_result.py -i organism/organism/

5. smile_list.py (takes the previous output and converts the ligands into their SMILES. It requires chembl_#.db for SQL)

  $ python3 smile_list.py -i organism/organism/ 

6. lt_csv_maker/lt_id_smiles/lt_smiles_maker.py (creates the final outputs by grouping the ligands by lt)

  $ python3 lt_csv_maker.py -i organism/organism/ -dict organism/organism_final.json (output from prepocessing_proteome_liqQ.py) -o organism_lt_ligands.csv

  $ python3 lt_id_smiles.py -i organism/organism/ -dict organism/organism_final.json (output from prepocessing_proteome_liqQ.py) -o organism_lt_id_smiles.csv

  $ python3 lt_smiles_maker.py -i organism/organism/ -dict organism/organism_final.json (output from prepocessing_proteome_liqQ.py) -o organism_lt_smiles.csv

7. clustering.py (Clusters the ligands by lt and searches for commercially available ligands in the Zinc database.)

Before running this last step, create in organism/organism/ the folder zinc/ ( $ mkdir organism/organism/zinc/ )

  $ python3 clustering.py -i organism/organism/organism_lt_id_smiles.csv -o organism/organism/zinc/

Test

We have a Test/ folder that contains the results of the complete script for the proteome of one of the dengue variants.