localpdb 0.1.0

Store, manage and query the local copy of the PDB (Protein Data Bank) resources.

localpdb

• Store a local copy of the PDB data and related protein resources,
• Access and query the data convinently through pandas DataFrame structures,
• Update with weekly releases of a new data with full history versioning,
• Customize to your needs or add other data sources with simple plugin system.

Quick start

• Install localpdb with pip and run setup script to download PDB data and protein structures in the PDB format:

pip3 install localpdb
localpdb_setup.py -db_path /path/to/your/localpdb --fetch_pdb

• Simple pipeline that selects for further analysis a representative set of protein structures:
  • solved with X-ray crystallography,
  • with resolution better than 2.5 angstroms,
  • deposited in 2010 or later,
  • with redundancy removed at the sequence level.

from localpdb import PDB
import gzip

lpdb = PDB(db_path='/path/to/your/localpdb')

# Select protein structures solved with X-ray diffraction (resolution above 2.5 A)
lpdb.entries = lpdb.entries.query('type == "prot"')
lpdb.entries = lpdb.entries.query('method == "diffraction"')
lpdb.entries = lpdb.entries.query('resolution <= 2.5')
lpdb.entries = lpdb.entries.query('deposition_date.dt.year >= 2010')

# Remove redundancy (select only representative structure from each sequence cluster)
lpdb.load_clustering_data(cutoff=90)
lpdb.chains = lpdb.chains[lpdb.chains['clust-90'].notnull()]

representative = lpdb.chains.groupby(by='clust-90')['resolution'].idxmin()
lpdb.chains = lpdb.chains.loc[representative]

for pdb_fn in lpdb.chains.pdb_fn:
    # your analysis here

Additional resources

(In development)

Plugins

(In development)