PKPDB

A database of over 12M theoretical pKa values calculated over 120k protein structures deposited in the Protein Data Bank. DOI: 10.1093/bioinformatics/btab518

While this repo contains all the necessary code to replicate pKPDB, you may download the full data from pypka.org/pKPDB.

@article{reis2021pKPDB,
    author = {Reis, Pedro B. P. S. and Clevert, Djork-Arné and Machuqueiro, Miguel},
    title = "{pKPDB: a Protein Data Bank extension database of pKa and pI theoretical values}",
    journal = {Bioinformatics},
    year = {2021},
    month = {07},
    issn = {1367-4803},
    doi = {10.1093/bioinformatics/btab518}
}

Create/Update Protein Data Bank proteins

dbs/init.sql contains the schema of the database

dbs/update_db.sh downloads the latest entries to the Protein Data Banks and inserts them into the database

cd initial/
psql -d pkpdb -f init.sql

cd PDB_data/
bash update_db.sh

mmseqs database of fasta files also needs to be updated

cd src/extra_properties/DB_PDB/
mmseqs databases PDB ./DB_PDB tmp
rm -rf tmp

Add proteins to the database

# Add a random protein
python3 fill.py

# Add the protein with the IDCODE <IDCODE>
python3 fill.py --idcode <IDCODE>

Extra options:

• ncpus NCPUS
• verbose {DEBUG,INFO,WARNING}
• nres-limit NRES_LIMIT

Dependencies

python3 -m pip install psycopg2-binary sqlalchemy pypka biopython python-decouple

mmseqs and DSSP are also required for running extra_properties/solvent_exposure.py