def _create_struct_dict(interval_dict, ss_dis, uni_folder):
""" Creates the structure dictionary.
Args:
interval_dict (dict): A dictionary in the following form:
{
'11BG_A_Q3E840':
[
[
[1, 124],
[27, 150]
]
]
}
uni_folder (Unicode): A path to the folder that has single
UniProt fasta files.
ss_dis: a dictionary extracted from ss_dis.txt, in the following form:
ss_dis[pdb_A] = {
'sequence': '',
'secstr': '',
'disorder': ''
}
Returns:
A dictionary in the following form:
{
'PDB_CHAIN': [],
'UNIPROT': [],
'SEC_STRUCT': []
}
"""
structure_dict = {'PDB_CHAIN': [], 'SP_PRIMARY': [], 'SEC_STRUCT': []}
for pdb_chain_uni in interval_dict:
pdb_chain = ''.join([
pdb_chain_uni.split('_')[0],
'_',
pdb_chain_uni.split('_')[1]
])
uni_id = pdb_chain_uni.split('_')[2]
uni_fp = build_abs_path(uni_folder, uni_id)
len_uni_seq = len(
(SeqIO.read(open(uni_fp), "fasta")).seq
)
disorder = ss_dis[pdb_chain]['disorder']
ss = ss_dis[pdb_chain]['secstr']
intervals = interval_dict[pdb_chain_uni]
pdb_struct = _create_pdb_struct(intervals, disorder, ss, len_uni_seq)
structure_dict['PDB_CHAIN'].append(pdb_chain)
structure_dict['SP_PRIMARY'].append(uni_id)
structure_dict['SEC_STRUCT'].append(pdb_struct)
return structure_dict