def MINT_homo_tenTrial_to_csvData():
predPPIs = {}
ggi_df, ppi_df = MINT.parse_MINT(
ppiFile='./data/MINT/species human',
uniProtMap="./data/UniProt/uniprot-taxonomy_9606.tab",
wFile_GGI='./data/parsed/MINT_homo_GGI.pkl',
wFile_PPI='./data/parsed/MINT_homo_PPI.pkl',
root='../')
dataset_len = int(int(len(ppi_df.index) * 0.5) * 0.1)
with open('./resultData/h**o/trimmed_predPPIs_homo.json', 'r') as f:
for line in f.readlines():
predPPIs.update(json.loads(line))
geneToEntry = helper.uniprot_map()
baseTags = [
'commonNeighbor', 'L3uvJoin', 'xyContrib_dualCN_uvJoin', 'CRA',
'CH2_L3', 'Sim'
]
for dataset in ["MINT_homo"]:
tags = [
"{}_tenTrial_{}".format(baseTag, dataset)
for baseTag in baseTags
]
for i in range(len(tags)):
tag = tags[i]
PPIsList = [[[geneToEntry[g[0]], geneToEntry[g[1]]]
for g in j[:dataset_len]
if g[0] in geneToEntry and g[1] in geneToEntry]
for j in predPPIs[tag]]
for j in range(len(PPIsList)):
with open('./GoSemSimPrepData/{}_{}.csv'.format(tag, j),
'w') as f:
f.write(
"\n".join(['nodeA\tnodeB'] +
["\t".join(ppi)
for ppi in PPIsList[j]]) + "\n")
def run_for_a_trajectory(PARMS, filename, nucleotides, charges):
universe = MDAnalysis.Universe(PARMS["file_name"], filename)
TimeTable = MINT.ReadInTrajectory(nucleotides, universe, PARMS)
PARMS["last_frame"] = len(universe.trajectory)
manager = multiprocessing.Manager()
trajs, num_of_frames, should = MINT.divide_trajectory(PARMS)
print "Trajectory ", filename, "has ", len(universe.trajectory),
name = filename.replace(".dcd", "")
print " frames, running for ", num_of_frames
que = manager.list()
ths = [
multiprocessing.Process(target=for_a_sub_traj,
args=(nucleotides, charges, PARMS, TimeTable,
trajs[i], que, name))
for i in range(PARMS["threads"])
]
for p in ths:
p.start()
for p in ths:
p.join()
return que
def run():
PARMS = MINT.inside_read_in_parms()
PARMS["OUT_FILE"] = open(
PARMS["working_dir"] + "/" + PARMS["out_name"] + "_hbonds_log.txt",
"w")
nucleotides = MINT.get_nucleic_from_pdb(PARMS)
charges = MINT.read_in_charges(nucleotides, PARMS)
pickles = {}
nuc_nums = []
if PARMS["nucleotides"] != "":
for i in PARMS["nucleotides"].split(";"):
if i:
tmp = [
a for a in i.replace("(", "").replace(")", "").split('-')
if a
]
nuc_nums.extend(range(int(tmp[0]), int(tmp[1]) + 1))
nuc = []
for i in nucleotides:
if i.get_id()[1] in nuc_nums:
nuc.append(i)
nucleotides = nuc
if "out_dictionaries_MINT" in PARMS["files_dcd"][0]:
read_in_mint_pickles(
PARMS["working_dir"] + "/" + PARMS["files_dcd"][0],
PARMS["working_dir"] + "/" + PARMS["out_name"] + ".csv")
else:
for filename in PARMS["files_dcd"]:
filename_and_dir = PARMS["working_dir"] + filename
pickles[filename] = run_for_a_trajectory(PARMS, filename_and_dir,
nucleotides, charges)
put_together(pickles,
PARMS["working_dir"] + PARMS["out_name"] + ".csv",
PARMS["files_dcd"], PARMS["working_dir"])
print "Written to file", PARMS["working_dir"] + PARMS[
"out_name"] + ".csv"
def for_a_sub_traj(nucleotides, charges, PARMS, TimeTable, ran, que, name):
out = {}
ppkl = name + "_" + str(min(ran)) + "_" + str(max(ran)) + ".pkl"
if os.path.isfile(ppkl) and PARMS["only_analysis"]:
que.append(ppkl)
print "Not running for ", ppkl
else:
print " running for ", ppkl
for N in ran:
dd = MINT.measure_for_all(nucleotides, charges, PARMS, TimeTable,
N)
out[N] = sum_of_hbonds(dd)
out[N].extend(sum_of_stacking(dd))
pickle.dump(out, open(ppkl, "wb"))
que.append(ppkl)
def append_precRecMap_multiCore(fNames,
predPPI,
samplePPI,
datasetClass,
coreNo,
isGGI=False,
logging=False):
if isGGI: i = 0
else: i = 1
fullPPISet = {
'bioGRID': [
list(ppi)
for ppi in np.asarray([*bg.parse_bioGRID(
root='../')][i][['nodeA', 'nodeB']])
],
'STRING': [
list(ppi)
for ppi in np.asarray([*string.parse_STRING(
root='../')][i][['nodeA', 'nodeB']])
],
'MINT': [
list(ppi) for ppi in np.asarray([*MINT.parse_MINT(
root='../')][i][['nodeA', 'nodeB']])
],
'IntAct_spoke': [
list(ppi) for ppi in np.asarray([
*IntAct.parse_IntAct(root='../', spokeModel=True)
][i][['nodeA', 'nodeB']])
]
}
fullPrecRecMap = {}
if not os.path.exists("./resultData/PRCurveMap.json"):
with open("./resultData/PRCurveMap.json", "w") as f:
f.write(json.dumps(fullPrecRecMap))
precRecMap = ppiLPred.precRecMap_multiCore(
fNames, predPPI, samplePPI, [fullPPISet[i] for i in datasetClass],
coreNo, logging)
with open('./resultData/PRCurveMap.json', 'r') as f:
fullPrecRecMap = json.loads(f.read())
fullPrecRecMap.update(precRecMap)
with open('./resultData/PRCurveMap.json', 'w') as f:
f.write(json.dumps(fullPrecRecMap))
def trim_ppi_result(fNames, datasetClass):
# get only the top PPI & scores equal to the size of its original dataset
trimNum = {
'bioGRID':
int(len([*bg.parse_bioGRID(root='../')][1].index) * 0.5),
'STRING':
int(len([*string.parse_STRING(root='../')][1].index) * 0.5),
'MINT':
int(len([*MINT.parse_MINT(root='../')][1].index) * 0.5),
'IntAct_spoke':
int(
len([*IntAct.parse_IntAct(root='../', spokeModel=True)
][1].index) * 0.5)
}
if not os.path.exists('./resultData/trimmed_predPPIs.json'):
with open('./resultData/trimmed_predPPIs.json', 'w') as f:
pass
with open('./resultData/trimmed_predScores.json', 'w') as f:
pass
for i in range(len(fNames)):
predPPI, predScore = [], []
with open("./resultData/{}_PPI.json".format(fNames[i]), 'r') as f:
for line in f.readlines():
predPPI.append(
json.loads(line)[0:trimNum[datasetClass[i]]])
with open("./resultData/{}_score.json".format(fNames[i]),
'r') as f:
for line in f.readlines():
predScore.append(
json.loads(line)[0:trimNum[datasetClass[i]]])
predPPIs, predScores = {fNames[i]: predPPI}, {fNames[i]: predScore}
with open('./resultData/trimmed_predPPIs.json', 'a+') as f:
f.write(json.dumps(predPPIs) + "\n")
with open('./resultData/trimmed_predScores.json', 'a+') as f:
f.write(json.dumps(predScores) + "\n")
def trim_multiple_ppi_result(fNames, datasetClass, trialSize):
trimNum = {
'bioGRID':
int(len([*bg.parse_bioGRID(root='../')][1].index) * 0.5),
'STRING':
int(len([*string.parse_STRING(root='../')][1].index) * 0.5),
'MINT':
int(len([*MINT.parse_MINT(root='../')][1].index) * 0.5),
'IntAct_spoke':
int(
len([*IntAct.parse_IntAct(root='../', spokeModel=True)
][1].index) * 0.5)
}
if not os.path.exists('./resultData/trimmed_predPPIs.json'):
with open('./resultData/trimmed_predPPIs.json', 'w') as f:
pass
with open('./resultData/trimmed_predScores.json', 'w') as f:
pass
for i in range(len(fNames)):
predPPI, predScore = [], []
for j in range(trialSize):
with open("./resultData/{}_{}_PPI.json".format(fNames[i], j),
'r') as f:
predPPI.append(
json.loads(f.read())[0:trimNum[datasetClass[i]]])
with open("./resultData/{}_{}_score.json".format(fNames[i], j),
'r') as f:
predScore.append(
json.loads(f.read())[0:trimNum[datasetClass[i]]])
predPPIs, predScores = {fNames[i]: predPPI}, {fNames[i]: predScore}
with open('./resultData/trimmed_predPPIs.json', 'a+') as f:
f.write(json.dumps(predPPIs) + "\n")
with open('./resultData/trimmed_predScores.json', 'a+') as f:
f.write(json.dumps(predScores) + "\n")
def trim_multiple_ppi_result(fNames, datasetClass, trialSize):
bioGRID_homo = int(
len([
*bg.parse_bioGRID(
filename=
'./data/BioGRID/BIOGRID-ORGANISM-Homo_sapiens-3.5.187.tab2.txt',
wFile_GGI='./data/parsed/BioGRID_homo_GGI.pkl',
wFile_PPI='./data/parsed/BioGRID_homo_PPI.pkl',
root="../")
][1].index) * 0.5)
STRING_homo = int(
len([
*string.parse_STRING(
ppiFile='./data/STRING/9606.protein.links.v11.0.txt',
typeFile='./data/STRING/9606.protein.actions.v11.0.txt',
uniProtMap=
'./data/UniProt/uniprot-taxonomy_9606_STRING.tab',
root='../',
wFile_GGI='./data/parsed/STRING_homo_GGI.pkl',
wFile_PPI='./data/parsed/STRING_homo_PPI.pkl')
][1].index) * 0.5)
MINT_homo = int(
len([
*MINT.parse_MINT(
ppiFile='./data/MINT/species human',
uniProtMap="./data/UniProt/uniprot-taxonomy_9606.tab",
wFile_GGI='./data/parsed/MINT_homo_GGI.pkl',
wFile_PPI='./data/parsed/MINT_homo_PPI.pkl',
root="../")
][1].index) * 0.5)
trimNum = {
'HuRI': int(len(HuRI.parse_HuRI(root='../').index) * 0.5),
"bioGRID_homo": bioGRID_homo,
"STRING_homo": STRING_homo,
"MINT_homo": MINT_homo
}
if not os.path.exists('./resultData/h**o/trimmed_predPPIs.json'):
with open('./resultData/h**o/trimmed_predPPIs.json', 'w') as f:
pass
with open('./resultData/h**o/trimmed_predScores.json', 'w') as f:
pass
for i in range(len(fNames)):
predPPI, predScore = [], []
for j in range(trialSize):
with open(
"./resultData/h**o/{}_{}_PPI.json".format(
fNames[i], j), 'r') as f:
predPPI.append(
json.loads(f.read())[0:trimNum[datasetClass[i]]])
with open(
"./resultData/h**o/{}_{}_score.json".format(
fNames[i], j), 'r') as f:
predScore.append(
json.loads(f.read())[0:trimNum[datasetClass[i]]])
predPPIs, predScores = {fNames[i]: predPPI}, {fNames[i]: predScore}
with open('./resultData/h**o/trimmed_predPPIs.json', 'a+') as f:
f.write(json.dumps(predPPIs) + "\n")
with open('./resultData/h**o/trimmed_predScores.json', 'a+') as f:
f.write(json.dumps(predScores) + "\n")
def append_precRecMap_multiCore(fNames,
predPPI,
samplePPI,
datasetClass,
coreNo,
isGGI=False,
logging=False):
if isGGI: i = 0
else: i = 1
bioGRID_homo = [
list(ppi) for ppi in np.asarray([
*bg.parse_bioGRID(
filename=
'./data/BioGRID/BIOGRID-ORGANISM-Homo_sapiens-3.5.187.tab2.txt',
wFile_GGI='./data/parsed/BioGRID_homo_GGI.pkl',
wFile_PPI='./data/parsed/BioGRID_homo_PPI.pkl',
root="../")
][i][['nodeA', 'nodeB']])
]
STRING_homo = [
list(ppi) for ppi in np.asarray([
*string.parse_STRING(
ppiFile='./data/STRING/9606.protein.links.v11.0.txt',
typeFile='./data/STRING/9606.protein.actions.v11.0.txt',
uniProtMap=
'./data/UniProt/uniprot-taxonomy_9606_STRING.tab',
root='../',
wFile_GGI='./data/parsed/STRING_homo_GGI.pkl',
wFile_PPI='./data/parsed/STRING_homo_PPI.pkl')
][i][['nodeA', 'nodeB']])
]
MINT_homo = [
list(ppi) for ppi in np.asarray([
*MINT.parse_MINT(
ppiFile='./data/MINT/species human',
uniProtMap="./data/UniProt/uniprot-taxonomy_9606.tab",
wFile_GGI='./data/parsed/MINT_homo_GGI.pkl',
wFile_PPI='./data/parsed/MINT_homo_PPI.pkl',
root="../")
][i][['nodeA', 'nodeB']])
]
fullPPISet = {
'HuRI': [
list(ppi) for ppi in np.asarray(
HuRI.parse_HuRI(root='../')[['nodeA', 'nodeB']])
],
'bioGRID_homo':
bioGRID_homo,
'STRING_homo':
STRING_homo,
'MINT_homo':
MINT_homo
}
fullPrecRecMap = {}
if not os.path.exists("./resultData/PRCurveMap_homo.json"):
with open("./resultData/PRCurveMap_homo.json", "w") as f:
f.write(json.dumps(fullPrecRecMap))
precRecMap = ppiLPred.precRecMap_multiCore(
fNames, predPPI, samplePPI, [fullPPISet[i] for i in datasetClass],
coreNo, logging)
with open('./resultData/PRCurveMap_homo.json', 'r') as f:
fullPrecRecMap = json.loads(f.read())
fullPrecRecMap.update(precRecMap)
with open('./resultData/PRCurveMap_homo.json', 'w') as f:
f.write(json.dumps(fullPrecRecMap))