def data237_fp_gen(refresh=False):
for complex_id in complex_ids(path = data237_complex_root):
output_path = os.path.join(data237_fp6_root,complex_id)
if not refresh and os.path.exists(output_path):
print("%s already processed" %complex_id)
continue
print "start processing %s" %complex_id
data_dir = os.path.join(data237_complex_root,complex_id)
antigen = load_pdb_struct(os.path.join(data_dir,"antigen.pdb"),residue_cls = Residue)
antibody = load_pdb_struct(os.path.join(data_dir,"antibody.pdb"),residue_cls = Residue)
try:
c = Complex(complex_id, antigen,antibody)
fps = c.gen_fps()
except:
sys.stderr.write("complex %s encountered error.\n" %complex_id)
if not os.path.exists(output_path):
os.makedirs(output_path)
for i,fp in enumerate(fps):
path = os.path.join(output_path,"fp%d.txt" %(i+1))
print "to path", path
fp.tofile(path)
def gen_dist_matrix(plane_type, atg_or_atb, res_or_tri):
data_dir = os.path.join(data237_fp175_padded_root, "%s-%s-%s" % (plane_type, atg_or_atb, res_or_tri))
dataloader = make_dataloader(data_dir, make_single_line_converter(None))
c_id_list = complex_ids(data_dir)
complex_pairwise_calc(c_id_list, dataloader, corr_coef, callback=callback)
def print_splitted_fp_in_csv(res_or_tri, which=0):
"""
split 360 bits into parts
"""
data_dir = os.path.join(data480_root, "fp_aaindex_if_padded" if res_or_tri == "res" else "fp_aaindex_if_padded_tri")
delimiter = ","
cids = complex_ids(data_dir)
name = ("%s 3 bits" % (res_or_tri), "atb 4 bits", "15 bits")
from ve.fp.complex_util.padding import PaddedComplexFingerPrint, OverallSpatialDistribution
atg_dist, atb_dist, tri_dist = OverallSpatialDistribution.from_cache()
first_3_count = sum(atg_dist.values() if res_or_tri == "res" else tri_dist.values()) * 3
second_3_count = sum(atb_dist.values()) * 3
last_15_count = sum(atg_dist.values()) * 15
first_3_dataloader = make_dataloader(data_dir, make_single_line_converter(slice(0, first_3_count)))
second_3_dataloader = make_dataloader(
data_dir, make_single_line_converter(slice(first_3_count, first_3_count + second_3_count))
)
last_15_dataloader = make_dataloader(
data_dir,
make_single_line_converter(
slice(first_3_count + second_3_count, first_3_count + second_3_count + last_15_count)
),
)
dls = [first_3_dataloader, second_3_dataloader, last_15_dataloader]
print ("cid, %s" % name[which])
for cid in sorted(cids):
print ("%s,%s" % (cid, delimiter.join(map(lambda d: "%.2f" % d, dls[which](cid)))))
def gen_pairwise_dist(res_or_tri):
data_dir = os.path.join(data480_root, "fp_aaindex_if_padded" if res_or_tri == "res" else "fp_aaindex_if_padded_tri")
dataloader = make_dataloader(data_dir, make_single_line_converter(None))
c_id_list = complex_ids(data_dir)
complex_pairwise_calc(c_id_list, dataloader, corr_coef, callback=callback)
def print_fp_in_csv(plane_type, atg_or_atb, res_or_tri):
data_dir = os.path.join(data237_fp175_padded_root, "%s-%s-%s" % (plane_type, atg_or_atb, res_or_tri))
dataloader = make_dataloader(data_dir, make_single_line_converter(None))
delimiter = ","
cids = complex_ids(data_dir)
for cid in sorted(cids):
print ("%s,%s," % (cid, delimiter.join(map(lambda d: "%.2f" % d, dataloader(cid)))))
def get_overall_residue_distribution():
from ve.util.load_pdb import load_complexes, complex_ids
from ve.fp.test.common import GeometryResidue
cs = load_complexes(complex_ids(), complex_cls=ComplexWithResidueSpatialDistribution, residue_cls=GeometryResidue)
overall_atg_dist, overall_atb_dist, overall_tri_dist = OverallSpatialDistribution.overall_dist(cs)
return overall_atg_dist, overall_atb_dist, overall_tri_dist
def setUp(self):
from ve.util.load_pdb import load_complexes, complex_ids
from ve.fp.fp_80 import Residue
from random import sample
sample_count = 10
sampled_ids = sample(complex_ids(), sample_count)
self.complexes = load_complexes(sampled_ids, complex_cls = ResiduePlaneBasedComplex, residue_cls = Residue)
def main():
from ve.util.complex import BaseComplex
from ve.fp.fp_80 import Residue as Residue80
class Complex(BaseComplex, ResidueTriangleTrait):
pass
from ve.util.load_pdb import load_complexes, complex_ids
for c in load_complexes(complex_ids(), complex_cls = Complex, residue_cls = Residue80):
c.get_triangles(refresh = True)
def main_480():
import os
from ve.util.load_pdb import complex_ids
from ve.config import data480_root
from data480 import fp370_atg_dataloader, fp370_atb_dataloader
fp_dir = os.path.join(data480_root, "fp_370_atg")
cids = complex_ids(fp_dir)
cids = map(lambda x: x.split(",")[0], cids)
complex_pairwise_calc(cids, fp370_atb_dataloader, corr_coef, callback = callback)
def simmat_from_db():
from ve.util.load_pdb import complex_ids
from ve.config import epi166_fp
from ve.dbconfig import db
cids = complex_ids(epi166_fp)
mat = lmatrix.from_db(db["epi_166"], cids)
mat = lmatrix(cids, data=mat / np.matrix(mat.diagonal()).T)
print mat.to_csv_str()
def main1():
"""
main function
generate the 370-bit finger prints for all complexes in 237 dataset
"""
from ve.util.load_pdb import complex_ids
c_ids = complex_ids()
for cid in c_ids:
print cid
gen_fp_for_complex(cid)
def simmat_from_db():
from ve.util.load_pdb import complex_ids
from ve.config import epi166_fp
from ve.dbconfig import db
cids = complex_ids(epi166_fp)
mat = lmatrix.from_db(db["epi_166"], cids)
mat = lmatrix(cids, data = mat / np.matrix(mat.diagonal()).T)
print mat.to_csv_str()
def test_count():
import os
from ve.util.load_pdb import complex_ids, load_complexes
from ve.config import data480_root, data480_complex_root
ids = complex_ids(data480_complex_root)
fp_dir = os.path.join(data480_root, "fp_370_atg" )
for cid in ids:
fp_path = os.path.join(fp_dir, "%s.fp" %cid)
if os.path.exists(fp_path):
print cid, "exists"
else:
print cid
def main(fp_dir, use_complex_plane = True, atg_as_rec = True, use_tri = True, use_cache = True):
from ve.fp.fp_80 import Residue
from ve.util.load_pdb import complex_ids, load_complexes
from ve.config import data237_fp175_padded_root, data480_complex_root
complex_cls = ComplexPlaneBasedComplex if use_complex_plane else ResiduePlaneBasedComplex
cids = complex_ids()
cs = load_complexes(cids, complex_cls = complex_cls, residue_cls = Residue)
for c in cs:
try:
fp_str = c.gen_fp_str(atg_as_receptor = atg_as_rec, use_cache = use_cache, use_tri = use_tri)
#print c.c_id
with open(fp_dir + "/" + c.c_id + ".fp", "w") as f:
f.write(fp_str)
except:
print "%s encountered error" %c.c_id
def main(fp_dir, res_or_tri= "tri", atg_as_rec = True, use_cache = True):
from ve.fp.fp_80 import Residue
from ve.util.load_pdb import complex_ids, load_complexes
from ve.config import data237_fp175_padded_root, data480_complex_root
cids = complex_ids()
cids = ["1FJ1_F", "3BN9_B","3B9K_EF"]
cs = load_complexes(cids, complex_cls = Complex, residue_cls = Residue)
for c in cs:
print c.c_id
fp_str = c.gen_fp_str(res_or_tri, atg_as_receptor = atg_as_rec, use_cache = use_cache)
try:
#fp_str = c.gen_fp_str(res_or_tri, atg_as_receptor = atg_as_rec, use_cache = use_cache)
with open(fp_dir + "/" + c.c_id + ".fp", "w") as f:
f.write(fp_str)
except:
print "%s encountered error" %c.c_id
def main2(which_as_rec):
import os
from ve.util.load_pdb import complex_ids, load_complexes
from ve.config import data480_root, data480_complex_root
ids = complex_ids(data480_complex_root)
cs = load_complexes(ids, directory = data480_complex_root,complex_cls = MyComplex, residue_cls = MyResidue)
fp_dir = os.path.join(data480_root, "fp_370_%s" %which_as_rec)
for c in cs:
fp_path = os.path.join(fp_dir, "%s.fp" %c.c_id)
if os.path.exists(fp_path):
print "%s preexists\n" %c.c_id
else:
print "processing", c.c_id
try:
fp = c.get_fp(which_as_rec)
fp.tofile(fp_path)
except Exception as e:
from ve.util.error import get_error_info
print c.c_id, get_error_info(e)
print c.c_id, "processed\n"
def main_237():
filter_list = set(["1UWX_AP"])
c_id_list = set(complex_ids()) - filter_list
# complex_pairwise_calc(c_id_list, src.fp370_atg_dataloader, corr_coef, callback = callback)
complex_pairwise_calc(c_id_list, src.fp370_atb_dataloader, corr_coef, callback = callback)
c = ComplexDual(complex_id, antigen, antibody)
c.gen_fp_to_file(complex_id, fp_type="double")
def already_processed(complex_id):
file_types = ["atg.fp","15bits.fp","atb.fp"]
calc_types = ["single", "double"]
for ct in calc_types:
for ft in file_types:
file_dir = os.path.join(data237_fps808015_root, ct, complex_id, ft)
print file_dir
if not os.path.exists(file_dir): return False#not processed completely
return True#processed completely
if __name__ == "__main__":
ids = ["1SLG_D", "1N4X_L", "1JV5_A", "1STS_B"]
for c_id in complex_ids(path = data237_complex_root):
if already_processed(c_id):
print "already processed", c_id
continue
else:
print "processing", c_id
try:
gen_fp(c_id)
except:
import sys
sys.stderr.write("error encountered processing %s" %c_id)
continue
if only_paratope:
self.find_epitope()
self.write_epitope(complex_dir, paraepi_dir)
if __name__ == "__main__":
import sys
from ve.util.load_pdb import complex_ids, load_complexes
from ve.fp.complex_util.paraepi import ParatopeNotFoundError, EpitopeNotFoundError
from ve.config import data_root
var = sys.argv[1]
complex_dir = os.path.join(data_root,"three-groups/split-complexes", var)
#get the complex ids
ids = complex_ids(complex_dir)
print ids
cs = load_complexes(ids, directory = complex_dir, complex_cls = ParaEpiGen, residue_cls = TestResidue)
for c in cs:
try:
c.gen_paraepi(complex_dir, os.path.join(data_root,"three-groups/paraepi", var), only_paratope = True)
except ParatopeNotFoundError:
print "paratope not found"
except EpitopeNotFoundError:
print "epitope not found"
