def main(argv=sys.argv):
if len(argv) != 3:
print "\n Usage: %s num_pairs.dat potential.dat" % argv[0]
print ""
sys.exit(1)
inf = open(argv[1], "r")
dc = pickle.load(inf)
inf.close()
p = dist.PTS(dc)
outf = open(argv[2], "w")
pickle.dump(p, outf)
outf.close()
def main():
file_list=GetFileList()
dc_fn='dist_splited_train_0525.dat'
dc=dist.UnformatedInput(dc_fn)
pts=dist.PTS(dc)
for i in range(len(file_list)):
lig_fn=file_list[i][0]
pro_fn=file_list[i][1]
pdbid=file_list[i][2]
if os.path.isfile(lig_fn):
lig=pybel.readfile('sdf',lig_fn).next()
pro=pybel.readfile('pdb',pro_fn).next()
score,scorev=GetFinger(lig,pro,pts,pdbid)
print score,scorev
def singleTest(dc_fn):
lig = pybel.readfile(
'sdf',
'E:/brd4/PMF_Yolanda/dataSet/version2/trainsetLigand/PDB_3U5L.sdf'
).next()
pro = pybel.readfile(
'pdb',
'E:/brd4/PMF_Yolanda/dataSet/version2/trainsetPocket/PDB_3U5L.pdb'
).next()
dc = dist.UnformatedInput(dc_fn)
pts = dist.PTS(dc)
dd = GetDscrptor(lig, pro, pts)
dscrptors = [str(dd[d]) for d in DESCRIPTOR]
for i in DESCRIPTOR:
print i, dd[i]
return dscrptors
def dscrptorOut(dc_fn):
dirs = {'Train':{'lig':'e:/brd4/PMF_Yolanda/dataSet/version4/trainsetLigand/',\
'pro':'e:/brd4/PMF_Yolanda/dataSet/version4/trainsetPocket/',\
'exp':'e:/brd4/PMF_Yolanda/dataSet/version4/dscrptorValueLE_train.txt',\
'exp_s':'e:/brd4/PMF_Yolanda/dataSet/version4/dscrptorValueLEScaled_train.txt'},\
'Test':{'lig':'e:/brd4/PMF_Yolanda/dataSet/version4/testsetLigand/',\
'pro':'e:/brd4/PMF_Yolanda/dataSet/version4/testsetPocket/',\
'exp':'e:/brd4/PMF_Yolanda/dataSet/version4/dscrptorValueLE_test.txt',\
'exp_s':'e:/brd4/PMF_Yolanda/dataSet/version4/dscrptorValueLEScaled_test.txt'}
}
dc = dist.UnformatedInput(dc_fn)
pts = dist.PTS(dc)
def writeDES(ligdir, prodir, expdir):
with open(expdir, 'w') as f:
f.write('\t'.join(['Name', 'Active'] + DESCRIPTOR) + '\n')
for num, p in enumerate(os.listdir(ligdir)):
print p, num
a = 0
lig = pybel.readfile('sdf', ligdir + p).next()
pro = pybel.readfile('pdb', prodir + p[:-4] + '.pdb').next()
if lig.data['Active'] == '1': a = 1
dd = GetDscrptor(lig, pro, pts)
dscrptors = [str(dd[d]) for d in DESCRIPTOR]
f.write(p[:-4].replace('ss', '/') + '\t' + str(a) + '\t' +
'\t'.join(dscrptors) + '\n')
writeDES(dirs['Train']['lig'], dirs['Train']['pro'], dirs['Train']['exp'])
writeDES(dirs['Test']['lig'], dirs['Test']['pro'], dirs['Test']['exp'])
scaler = getScaler_fromFile(dirs['Train']['exp'])
table_train = scaleDES_fromFile(scaler, dirs['Train']['exp'])
table_test = scaleDES_fromFile(scaler, dirs['Test']['exp'])
with open(dirs['Train']['exp_s'], 'w') as ftrain:
ftrain.write('\t'.join(['Name', 'Active'] + DESCRIPTOR) + '\n')
for i in table_train:
ftrain.write('\t'.join(map(str, i)) + '\n')
with open(dirs['Test']['exp_s'], 'w') as ftest:
ftest.write('\t'.join(['Name', 'Active'] + DESCRIPTOR) + '\n')
for i in table_test:
ftest.write('\t'.join(map(str, i)) + '\n')
def dscrptorOut_robscale(dc_fn):
dirs = {'Train':{'lig':'e:/brd4/PMF_Yolanda/dataSet/version3/trainsetLigand/',\
'pro':'e:/brd4/PMF_Yolanda/dataSet/version3/trainsetPocket/',\
'exp':'e:/brd4/PMF_Yolanda/dataSet/version3/dscrptorValueSscaled_train.txt'},\
'Test':{'lig':'e:/brd4/PMF_Yolanda/dataSet/version3/testsetLigand/',\
'pro':'e:/brd4/PMF_Yolanda/dataSet/version3/testsetPocket/',\
'exp':'e:/brd4/PMF_Yolanda/dataSet/version3/dscrptorValueSscaled_test.txt'}
}
trainScaler = None
dc = dist.UnformatedInput(dc_fn)
pts = dist.PTS(dc)
def tableGen(ligdir, prodir):
dscrptorTable, activTable = [], []
for num, p in enumerate(os.listdir(ligdir)):
print p, num
a = 0
lig = pybel.readfile('sdf', ligdir + p).next()
pro = pybel.readfile('pdb', prodir + p[:-4] + '.pdb').next()
if lig.data['Active'] == '1': a = 1
activTable.append([p[:-4].replace('ss', '/'), str(a)])
dd = GetDscrptor(lig, pro, pts)
dscrptorTable.append([dd[d] for d in DESCRIPTOR])
return dscrptorTable, activTable
def tableOut(dscTable, acTable, expdir):
with open(expdir, 'w') as f:
f.write('\t'.join(['Name', 'Active'] + DESCRIPTOR) + '\n')
for ac, dsc in zip(dscTable, acTable):
entry = ac + dsc
f.write('\t'.join([str(i) for i in entry]) + '\n')
dscTabTrain, acTabTrain = tableGen(dirs['Train']['lig'],
dirs['Train']['pro'])
dscTabTrain_s, trainScaler = scaleDES(dscrptorTable, trainScaler)
tableOut(dscTabTrain_s, acTabTrain, dirs['Train']['exp'])
dscTabTest, acTabTest = tableGen(dirs['Test']['lig'], dirs['Test']['pro'])
dscTabTest_s, trainScaler = scaleDES(dscrptorTable, trainScaler)
tableOut(dscTabTest_s, acTabTest, dirs['Test']['exp'])
def main(argv=sys.argv):
if len(argv) < 3:
print "\n Usage: %s outfile in.num[...]" % argv[0]
print " in.num: generated by pmf_atom_pairs.py"
print ""
sys.exit(1)
if os.path.exists(argv[1]):
print "Warning: %s already exists. Are you sure you've specified {outfile}?" % argv[
1]
sys.exit(1)
dir = os.path.dirname(argv[0])
dat_name = os.path.join(dir, "dist_splited_train_0525.dat")
dc = dist.UnformatedInput(dat_name)
global pts
pts = dist.PTS(dc)
outf = open(argv[1], "w")
print >> outf, "name\tPMF.score"
for name in argv[2:]:
for each_name in glob(name):
do_for_each(each_name, outf)
outf.close()
def main(argv=sys.argv):
if len(argv) != 2:
print "\n Usage: %s in.list"%argv[0]
print " in.list: each line should be 'proname ligname'"
print ""
sys.exit(1)
file_list=GetFileList(argv[1])
#to get atom pair potentional A_ij(r)
dirname = os.path.dirname(argv[0])
dc_fn = os.path.join(dirname,'dist_splited_train_0525.dat')
dc=dist.UnformatedInput(dc_fn)
pts=dist.PTS(dc)
for i in range(len(file_list)):
lig_fn=file_list[i][0]
pro_fn=file_list[i][1]
pdbid=file_list[i][2]
lig_format = lig_fn[lig_fn.rfind(".")+1:]
pro_format = pro_fn[pro_fn.rfind(".")+1:]
if os.path.isfile(lig_fn):
lig=pybel.readfile(lig_format,lig_fn).next()
pro=pybel.readfile(pro_format,pro_fn).next()
score,scorev = GetFinger(lig,pro,pts,pdbid)
print pdbid,score,scorev
def molGen(sdf, proString, pts, model, scaler):
with open(sdf) as sdfile:
molString = ''
for line in sdfile:
molString += line
if line.strip() == '$$$$':
yield (molString, proString, pts, model, scaler)
molString = ''
if __name__ == '__main__':
proString = open(pr).read()
pts = dist.PTS(dist.UnformatedInput(dc_fn))
model = loadModel(modelfile)
f, fsdout = open(expdir, 'w'), open(expsdf, 'w')
f.write('\t'.join(['Name']+DESCRIPTOR+['Prob'])+'\n')
poo = mul.Pool(2)
import time
t = time.time()
scaler = rdg.getScaler_fromFile(scalerFit)
for n,v in enumerate(poo.imap(singleRun, molGen(sdf,proString,pts,model,scaler))):
entry, molString = v
f.write('\t'.join(entry)+'\n')
fsdout.write(molString)
if n%100==0: print n, molString[:15], entry[-1], time.time()-t
print time.time()-t
import dist
import pmf_atom_typer as pat
import raw_print_dist as rpd
import resiDescriptorGen_Scaled_LE as rdg
import extract_pdb1 as ep
from pybel import *
import os
from math import exp
import mpi4py.MPI as MPI
import time
sdfile = '../'
DESCRIPTOR = rdg.DESCRIPTOR
proString = open('4GPJ_confPred.pdb').read()
pts = dist.PTS(dist.UnformatedInput('dist_scPDB_train_4.dat'))
scaler = rdg.getScaler_fromFile('dscrptorValueLE_train.txt')
def loadModel(modelfile):
rawPara = [float(l.strip().split()[1]) for l in open(modelfile)]
rawDES = [l.strip().split()[0] for l in open(modelfile)]
para = rawPara[:-1]
descriptor = rawDES[:-1]
intercept = rawPara[-1]
return descriptor, para, intercept
model = loadModel('./model160126.txt')
E = exp(1.0)