def _seed_alignment(self):
# Build a Seed Alignment
from hpf.pdb import get_seq, get_crystal
#kdrew: this gets the atom record sequence and not the full sequence
pdb_seq = get_seq(get_crystal(self._pdbid))
#print "_seed_alignment pdb sequence: ", pdb_seq
from hpf.amnh.align import SeedAlignmentFactory
from Bio.SeqRecord import SeqRecord
#kdrew: might change this to be pairwise2 alignment
factory = SeedAlignmentFactory()
alignment = factory.create(self._alignment,
[SeqRecord(pdb_seq,self._pdbid+"_atom_record")])
return alignment
def buildPDBIndex(self, domain):
# Now we have to build the mapper and index the columns it covers
# This maps protein residues to the PDB structure record.
from sqlalchemy.orm import object_session
from hpf.hddb.db import PDBSeqRes
parent_id = int(domain.parent_id[3:])
session = object_session(domain)
pdbseqres = session.query(PDBSeqRes).filter(
PDBSeqRes.sequence_key == parent_id).first()
protein_atom_mapper = ProteinToPDBAtom(self.protein,
domain,
self.alignment,
pdbseqres=pdbseqres)
# This maps protein residues to domain residues
protein_domain_mapper = ProteinDomainMapper(self.domain)
# This maps protein residues to alignment column numbers
protein_alignment_mapper = ProteinAlignmentMapper(
self.protein, self.alignment)
# Simple structure object to use for indices
structure = Structure(protein_atom_mapper._pdbid[0:4],
chain=protein_atom_mapper._pdbid[4])
print >> sys.stderr, "Protein", structure.id, structure.chain, protein_atom_mapper._pdbid
# Get the PDB object downloaded and parsed for the mapping.
chain = get_crystal(protein_atom_mapper._pdbid, structure=True)
# Run DSSP on the set if not already run.
dssp = self.index.getDssp(protein_atom_mapper._pdbid)
if dssp != None:
print >> sys.stderr, "Using dssp", protein_atom_mapper._pdbid
else:
print >> sys.stderr, "Running dssp", protein_atom_mapper._pdbid
features = DSSPFeatureSet(protein_atom_mapper, chain, dssp=dssp)
self.index.setDssp(protein_atom_mapper._pdbid, features.dssp)
# Iterate over the DSSP features and index them.
for protein_res, xtra in features:
column = protein_alignment_mapper[protein_res]
domain_res = protein_domain_mapper[protein_res]
structure_res = protein_atom_mapper[protein_res]
attribute = ColumnStructureAttribute(column, self.protein,
protein_res, self.domain,
domain_res, structure,
structure_res, xtra)
# This will index all of the attributes for later lookup and serialization.
self.index.add(attribute)
Yields protein indices paired with DSSP scores.
@return generator of (protein,chain[protein].xtra[key]) tuples.
"""
atom_res = list(self.chain)
for protein, structure in self.mapper:
if protein == None or structure == None:
continue
xtra = atom_res[structure].xtra
yield (protein, xtra)
if __name__ == "__main__":
pdbid = "1n6j"
from hpf.pdb import get_crystal
pdb = get_crystal("1yelA", structure=True, pdir="/tmp")
dssp = DSSP(pdb)
print "done"
print pdb
for res in pdb:
rasa = res.xtra[RASA] if res.xtra.has_key(RASA) else None
print res.get_resname()[1], rasa
from hpf.hddb.db import Session, Structure
session = Session()
pdb = list(session.query(Structure).get(1054611).pdb[0])[0]
dssp = DSSP(pdb)
for res in pdb:
rasa = res.xtra[RASA] if res.xtra.has_key(RASA) else None
print res.get_resname()[1], rasa
