def model_hsps(seq_id,
work_dir,
hsps,
refinement=REFINEMENT,
models_to_generate=MODELS_TO_GENERATE,
assessments=ASSESMENTS,
entries={},
pdb_divided="/data/databases/pdb/divided/",
tmp_dir=None,
max_models=3):
result = {"models": defaultdict(lambda: {})}
alns = []
for hsp in hsps:
aln = Struct(aln_query=Struct(name=seq_id,
seq=str(hsp.aln[0].seq),
start=hsp.query_start,
end=hsp.query_end),
aln_hit=Struct(name=hsp.hit.id,
seq=str(hsp.aln[1].seq),
start=hsp.hit_start,
end=hsp.hit_end))
alns.append(aln)
modeler = Modeller(work_dir, tmp_dir)
modeler._refinement = refinement
modeler.model_count = models_to_generate
modeler._assess_methods = assessments
modeler.parallel_jobs = 1
def pdb_fn(x):
return x.aln_hit.name.split("_")[0]
alns = sorted(alns,
key=lambda x: entries[pdb_fn(x)]
if pdb_fn(x) in entries else 20)
result["alns"] = alns
for aligment in alns[0:max_models]:
# pdb,aligment = pdb_alignment
pdb, chain, _, _ = aligment.aln_hit.name.split("_")
if not os.path.exists(
modeler.pdb_path(seq_id + "_" + aligment.aln_hit.name,
seq_id)):
base_model_path = pdb_divided + pdb[1:3] + "/pdb" + pdb + ".ent"
ChainSplitter(tmp_dir).make_pdb(base_model_path,
pdb,
chain,
overwrite=True)
models = modeler.create_model(
seq_id + "_" + aligment.aln_hit.name, aligment)
else:
models = [
modeler.pdb_path(seq_id + "_" + aligment.aln_hit.name,
seq_id, idx)
for idx in range(1, models_to_generate + 1)
]
result["models"][aligment.aln_hit.name] = models
return result
def run_dssp(self):
out = tempfile.mkstemp(suffix=".dssp")[1]
execute("dssp -i {pdb_path} -o {out}", pdb_path=self.pdb_path, out=out)
with open(out) as h:
start = False
for l in h:
if start:
res = int(l[5:10])
aa = l[10:14].strip()
ss = l[14:17].strip()
bbl1 = l[23:24]
bbl2 = l[24:25]
bp1 = int(l[25:29])
bp2 = int(l[29:33])
bslabel = l[33:34]
self.dssp.append(
Struct(res=res,
aa=aa,
ss=ss,
bp1=bp1,
bp2=bp2,
bbl1=bbl1,
bbl2=bbl2,
bslabel=bslabel))
else:
if l.startswith(" # RESIDUE AA"):
start = True
def residues_near_drug(drug_centroid, aa_residues):
residues_near = []
for r in aa_residues:
for a in list(r):
dist = a - Struct(coord=drug_centroid)
if dist > 20:
break
if dist < 10:
residues_near.append(r)
break
return residues_near
def query(self):
"hmmscan [-options] <hmmdb> <seqfile>"
self._format_database()
self._check_input_file()
if (not os.path.exists(self.output_file)) or (os.path.getsize(self.output_file) == 0):
self._hmmscan()
decorated = self.query_iterator()
meta = self.query_iterator()._meta
return Struct(_meta=meta, __iter__=lambda: decorated)
def smart_parse(path, seqs=None, gz_input=None):
"""
:param path: sequence path
:param seqs: dictionary of sequences. key=sequence name, value=Bio.Seq.Seq object
:return: sequences iterator
"""
raw_path = path.strip()
if gz_input or raw_path.endswith(".gz"):
path = path[:-3]
handle = gzip.open(raw_path, "rt")
else:
path = raw_path
handle = open(path, "r")
try:
it = None
if path.endswith(".fasta"):
it = bpio.parse(handle, "fasta")
if path.endswith(".faa"):
it = bpio.parse(handle, "fasta")
if path.endswith(".fna"):
it = bpio.parse(handle, "fasta")
if path.endswith(".gb"):
it = bpio.parse(handle, "gb")
if path.endswith(".gbf"):
it = bpio.parse(handle, "gb")
if path.endswith(".gbk"):
it = bpio.parse(handle, "gb")
if path.endswith(".genebank"):
it = bpio.parse(path, "gb")
if path.endswith(".gbff"):
it = bpio.parse(handle, "gb")
if path.endswith(".embl"):
it = bpio.parse(handle, "embl")
if path.endswith(".gff") or path.endswith(".gff3"):
from BCBio import GFF
it = GFF.parse(handle)
if path.endswith(".fq"):
it = bpio.parse(handle, "fastq")
if path.endswith(".fastq"):
it = bpio.parse(handle, "fastq")
if path.endswith(".hmm"):
it = bpsio.parse(handle, "hmmer3-text")
if path.endswith(".xml"):
with open(path) as h:
h.readline()
l = h.readline()
if "BlastOutput" in l:
it = add_blast_xml_props(
search_iterator(bpsio.parse(handle, "blast-xml")))
if "<uniprot" in l:
it = bpio.parse(handle, "uniprot-xml")
except:
handle.close()
if it:
if seqs:
def witer():
for x in it:
if x.id in seqs:
x.seq = seqs[x.id]
yield x
return Struct(__iter__=witer)
else:
return it
raise Exception("invalid format")
os.makedirs(self.model_directory(model_id, query_id))
if __name__ == '__main__':
from SNDG import init_log, Struct
init_log()
workdir = "/media/eze/Data/data/organismos/Pext14-3B/analysis/struct/good"
modeler = Modeller(workdir, "/tmp")
model_id = "PE143B_RS25640_3u52_B_6_498"
alignment = Struct(
aln_query=Struct(
name="PE143B_RS25640",
seq=
"KKLNAKDKYRLLTRDLAWEPSYRTEEEIFPYIAYEGLKIHDWNKWEDPFRLTMDAYWKYQAEKERKFYAIIDAHAQNNGHLNITDARYLSALKIFLQAISPGEYAAHKGFARAGREFRGVGTQVACQMQAIDELRHAQTQIHALSNYNKFYNGFHAFADQRDRIWYTSVARSFFDDAMSAGPFEFMIAIGFSFEYVLTNLLFVPFMSGAAYNGDMATVTFGFSAQSDEARHMTLGLECIKFMLEQDPANLPIVQGWIDKWFWRGFRVLGLVSTMMDYMLPKRVMSWREAWEIYGAENGGALFKDLARYGIRPPKSWDDAEASIDHMSHQFMLGLYQWSFGTAFHAWIPSDDDMQWLSAKYPTTFDKYYRPRWEHIKKMEAAGTPFKNYGLAKLCQCCQLPTVFTEPDDPTLICHRQVQYKGDKYHFCSDHCMGIFNNEPEKYIQAWLPMPALFQAPTN-GDLGAWMD-WVSLKDGQDNGDFADSQDRRN",
start=7,
end=494), # .ungap("-")
aln_hit=Struct(
name="3u52_B_6_498",
seq=
"KKLNLKDKYQYLTRDMAWEPTYQDKKDIFPEEDFEGIKITDWSQWEDPFRLTMDAYWKYQAEKEKKLYAIFDAFAQNNGHQNISDARYVNALKLFISGISPLEHAAFQGYSKVGRQFSGAGARVACQMQAIDELRHSQTQQHAMSHYNKHFNGLHDGPHMHDRVWYLSVPKSFFDDARSAGPFEFLTAISFSFEYVLTNLLFVPFMSGAAYNGDMATVTFGFSAQSDEARHMTLGLEVIKFILEQHEDNVPIVQRWIDKWFWRGFRLLSLVSMMMDYMLPNKVMSWSEAWEVYYEQNGGALFKDLERYGIRPPKYQDVANDAKHHLSHQLWTTFYQYCQATNFHTWIPEKEEMDWMSEKYPDTFDKYYRPRYEYLAKEAAAGRRFYNNTLPQLCQVCQIPTIFTEKDAPTMLSHRQIEHEGERYHFCSDGCCDIFKHEPEKYIQAWLPVHQIYQGNCEGGDLETVVQKYYHINIGEDNFDYVGSPDQKH",
start=0,
end=489))
ChainSplitter("/tmp/").make_pdb(
pdb_path="/data/databases/pdb/divided/u5/pdb3u52.ent",
pdb_id="3u52",
chain="B",
overwrite=True)
models = modeler.create_model(model_id, alignment)
def model_hsps(seq_id,
work_dir,
hsps,
refinement=REFINEMENT,
models_to_generate=MODELS_TO_GENERATE,
assessments=ASSESMENTS,
entries={},
tmp_dir=None,
max_models=3):
result = {"models": defaultdict(lambda: {}), "errors": []}
alns = []
for hsp in hsps:
aln = Struct(aln_query=Struct(name=seq_id,
seq=str(hsp.aln[0].seq),
start=hsp.query_start,
end=hsp.query_end),
aln_hit=Struct(name=hsp.hit.id,
seq=str(hsp.aln[1].seq),
start=hsp.hit_start,
end=hsp.hit_end))
alns.append(aln)
modeler = Modeller(work_dir, tmp_dir)
modeler._refinement = refinement
modeler.model_count = models_to_generate
modeler._assess_methods = assessments
modeler.parallel_jobs = 1
def pdb_fn(x):
return x.aln_hit.name.split("_")[0]
alns = sorted(alns,
key=lambda x: entries[pdb_fn(x)]
if pdb_fn(x) in entries else 20)
result["alns"] = alns
for aligment in alns[0:max_models]:
# pdb,aligment = pdb_alignment
if ";" in aligment.aln_hit.name:
pdb = aligment.aln_hit.name[-5:-1]
chain = aligment.aln_hit.name[-1]
modeller_pdb_id = f'{seq_id}_{pdb}_{chain}'
aligment.aln_hit.name = f'{pdb}_{chain}'
else:
pdb, chain, _, _ = aligment.aln_hit.name.split("_")
modeller_pdb_id = f'{seq_id}_{pdb}_{chain}'
if not os.path.exists(modeler.pdb_path(modeller_pdb_id, seq_id)):
try:
models = modeler.create_model(modeller_pdb_id, aligment)
except SequenceMismatchError as ex:
result["errors"].append(f"modeller_pdb_id {str(ex)}\n")
continue
except ModellerError as ex:
result["errors"].append(f"modeller_pdb_id {str(ex)}\n")
continue
else:
models = [
modeler.pdb_path(modeller_pdb_id, seq_id, idx)
for idx in range(1, models_to_generate + 1)
]
result["models"][aligment.aln_hit.name] = models
result["models"] = dict(result["models"])
return result