forked from jmlingeman/Network-Inference-Workspace
/
Helpers.py
220 lines (190 loc) · 10.6 KB
/
Helpers.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
from Network import *
import numpy
def predict_timepoint(ts_data, network, idx_to_use, idx_to_predict):
timepoint = ts_data[0].experiments[idx_to_use]
predict = ts_data[0].experiments[idx_to_predict]
diffs = []
for gene in ts_data[0].gene_list:
diffs.append(predict.ratios[gene] - timepoint.ratios[gene])
# Each prediction is a linear combination of
correct = 0
incorrect = 0
for i, gene1 in enumerate(network.gene_list):
predicted_value = 0
target_value = predict.ratios[gene1]
for gene2 in network.gene_list:
predicted_value += timepoint.ratios[gene1] * network.network[gene1][gene2]
if predicted_value > 0 and diffs[i] > 0:
correct += 1
elif predicted_value < 0 and diffs[i] < 0:
correct += 1
elif predicted_value == 0 and diffs[i] == 0:
correct += 1
else:
incorrect += 1
print "PERCENT CORRECT"
print float(correct) / len(network.gene_list)
print correct, incorrect, len(network.gene_list)
def generate_random_network(gene_list):
mat = numpy.random.randn(len(gene_list), len(gene_list))
mat = mat / numpy.sqrt(len(gene_list))
net = Network()
net.read_netmatrix(mat.tolist(), gene_list, True)
return net
def apply_dex(dex_gene, dex_storage, target_network, op="add"):
# Normalize the dex storage, take the bottom and top 20%
#dex_storage.normalize()
mod_network = target_network.copy()
#if op == "add":
#for gene in target_network.gene_list:
#print dex_storage.experiments[0].ratios[gene]
#print mod_network.network[gene][dex_gene]
#print mod_network.network[dex_gene][gene]
#mod_network.network[gene][dex_gene] += float(dex_storage.experiments[0].ratios[gene])
#mod_network.network[dex_gene][gene] += float(dex_storage.experiments[0].ratios[gene])
#if op == "mult":
print dex_storage.experiments[0].ratios
for gene in target_network.gene_list:
if dex_storage.experiments[0].ratios[gene] >= 2.0:
#mod_network.network[gene][dex_gene] *= float(dex_storage.experiments[0].ratios[gene])
mod_network.network[gene][dex_gene] = 1
elif dex_storage.experiments[0].ratios[gene] <= 0.5:
#mod_network.network[dex_gene][gene] *= -(1.0 / float(dex_storage.experiments[0].ratios[gene]))
mod_network.network[dex_gene][gene] = -1
else:
mod_network.network[dex_gene][gene] = 0.0001
return mod_network
def read_dko_index(filename):
file = open(file, 'r')
ind = []
for row in file:
ind.append(map(int, Set(row.strip().split)))
return ind
def get_example_data_files(name, settings):
# Read in gold standard network
goldnet = Network()
dko_file = None
if name == "small":
goldnet.read_goldstd(settings["global"]["small_network_goldnet_file"])
#Get a list of the knockout files
ko_file = settings["global"]["small_network_knockout_file"].split()
kd_file = settings["global"]["small_network_knockdown_file"].split()
ts_file = settings["global"]["small_network_timeseries_file"].split()
wt_file = settings["global"]["small_network_wildtype_file"].split()
mf_file = settings["global"]["small_network_multifactorial_file"].split()
elif name == "medium":
goldnet.read_goldstd(settings["global"]["medium_network_goldnet_file"])
#Get a list of the knockout files
ko_file = settings["global"]["medium_network_knockout_file"].split()
kd_file = settings["global"]["medium_network_knockdown_file"].split()
ts_file = settings["global"]["medium_network_timeseries_file"].split()
wt_file = settings["global"]["medium_network_wildtype_file"].split()
mf_file = settings["global"]["medium_network_multifactorial_file"].split()
elif name == "medium_2":
goldnet.read_goldstd(settings["global"]["medium2_network_goldnet_file"])
#Get a list of the knockout files
ko_file = settings["global"]["medium2_network_knockout_file"].split()
kd_file = settings["global"]["medium2_network_knockdown_file"].split()
ts_file = settings["global"]["medium2_network_timeseries_file"].split()
wt_file = settings["global"]["medium2_network_wildtype_file"].split()
mf_file = settings["global"]["medium2_network_multifactorial_file"].split()
elif name == "dream410":
goldnet.read_goldstd(settings["global"]["dream410_network_goldnet_file"])
#Get a list of the knockout files
ko_file = settings["global"]["dream410_network_knockout_file"].split()
kd_file = settings["global"]["dream410_network_knockdown_file"].split()
ts_file = settings["global"]["dream410_network_timeseries_file"].split()
wt_file = settings["global"]["dream410_network_wildtype_file"].split()
mf_file = settings["global"]["dream410_network_multifactorial_file"].split()
dko_file = settings["global"]["dream410_network_doubleknockout_file"].split()
dko_idx_file = settings["global"]["dream410_network_doubleknockout_index_file"].split()
elif name == "dream410_2":
goldnet.read_goldstd(settings["global"]["dream410_2_network_goldnet_file"])
#Get a list of the knockout files
ko_file = settings["global"]["dream410_2_network_knockout_file"].split()
kd_file = settings["global"]["dream410_2_network_knockdown_file"].split()
ts_file = settings["global"]["dream410_2_network_timeseries_file"].split()
wt_file = settings["global"]["dream410_2_network_wildtype_file"].split()
mf_file = settings["global"]["dream410_2_network_multifactorial_file"].split()
elif name == "dream410_3":
goldnet.read_goldstd(settings["global"]["dream410_3_network_goldnet_file"])
#Get a list of the knockout files
ko_file = settings["global"]["dream410_3_network_knockout_file"].split()
kd_file = settings["global"]["dream410_3_network_knockdown_file"].split()
ts_file = settings["global"]["dream410_3_network_timeseries_file"].split()
wt_file = settings["global"]["dream410_3_network_wildtype_file"].split()
mf_file = settings["global"]["dream410_3_network_multifactorial_file"].split()
elif name == "dream410_4":
goldnet.read_goldstd(settings["global"]["dream410_4_network_goldnet_file"])
#Get a list of the knockout files
ko_file = settings["global"]["dream410_4_network_knockout_file"].split()
kd_file = settings["global"]["dream410_4_network_knockdown_file"].split()
ts_file = settings["global"]["dream410_4_network_timeseries_file"].split()
wt_file = settings["global"]["dream410_4_network_wildtype_file"].split()
mf_file = settings["global"]["dream410_4_network_multifactorial_file"].split()
elif name == "dream410_5":
goldnet.read_goldstd(settings["global"]["dream410_5_network_goldnet_file"])
#Get a list of the knockout files
ko_file = settings["global"]["dream410_5_network_knockout_file"].split()
kd_file = settings["global"]["dream410_5_network_knockdown_file"].split()
ts_file = settings["global"]["dream410_5_network_timeseries_file"].split()
wt_file = settings["global"]["dream410_5_network_wildtype_file"].split()
mf_file = settings["global"]["dream410_5_network_multifactorial_file"].split()
elif name == "dream4100":
goldnet.read_goldstd(settings["global"]["dream4100_network_goldnet_file"])
#Get a list of the knockout files
ko_file = settings["global"]["dream4100_network_knockout_file"].split()
kd_file = settings["global"]["dream4100_network_knockdown_file"].split()
ts_file = settings["global"]["dream4100_network_timeseries_file"].split()
wt_file = settings["global"]["dream4100_network_wildtype_file"].split()
mf_file = settings["global"]["dream4100_network_multifactorial_file"].split()
elif name == "dream4100_2":
goldnet.read_goldstd(settings["global"]["dream4100_2_network_goldnet_file"])
#Get a list of the knockout files
ko_file = settings["global"]["dream4100_2_network_knockout_file"].split()
kd_file = settings["global"]["dream4100_2_network_knockdown_file"].split()
ts_file = settings["global"]["dream4100_2_network_timeseries_file"].split()
wt_file = settings["global"]["dream4100_2_network_wildtype_file"].split()
mf_file = settings["global"]["dream4100_2_network_multifactorial_file"].split()
elif name == "dream4100_3":
goldnet.read_goldstd(settings["global"]["dream4100_3_network_goldnet_file"])
#Get a list of the knockout files
ko_file = settings["global"]["dream4100_3_network_knockout_file"].split()
kd_file = settings["global"]["dream4100_3_network_knockdown_file"].split()
ts_file = settings["global"]["dream4100_3_network_timeseries_file"].split()
wt_file = settings["global"]["dream4100_3_network_wildtype_file"].split()
mf_file = settings["global"]["dream4100_3_network_multifactorial_file"].split()
elif name == "dream4100_4":
goldnet.read_goldstd(settings["global"]["dream4100_4_network_goldnet_file"])
#Get a list of the knockout files
ko_file = settings["global"]["dream4100_4_network_knockout_file"].split()
kd_file = settings["global"]["dream4100_4_network_knockdown_file"].split()
ts_file = settings["global"]["dream4100_4_network_timeseries_file"].split()
wt_file = settings["global"]["dream4100_4_network_wildtype_file"].split()
mf_file = settings["global"]["dream4100_4_network_multifactorial_file"].split()
elif name == "dream4100_5":
goldnet.read_goldstd(settings["global"]["dream4100_5_network_goldnet_file"])
#Get a list of the knockout files
ko_file = settings["global"]["dream4100_5_network_knockout_file"].split()
kd_file = settings["global"]["dream4100_5_network_knockdown_file"].split()
ts_file = settings["global"]["dream4100_5_network_timeseries_file"].split()
wt_file = settings["global"]["dream4100_5_network_wildtype_file"].split()
mf_file = settings["global"]["dream4100_5_network_multifactorial_file"].split()
elif name == "ecoli400":
goldnet.read_goldstd(settings["global"]["ecoli400_goldnet_file"])
#Get a list of the knockout files
ko_file = settings["global"]["ecoli400_knockout_file"].split()
kd_file = settings["global"]["ecoli400_knockdown_file"].split()
ts_file = settings["global"]["ecoli400_timeseries_file"].split()
wt_file = settings["global"]["ecoli400_wildtype_file"].split()
mf_file = settings["global"]["ecoli400_multifactorial_file"].split()
else:
print "ERROR: Dataset {0} not found, exiting.".format(name)
exit()
if dko_file != None:
return ko_file, kd_file, ts_file, wt_file, mf_file, goldnet, dko_file, dko_idx_file
else:
return ko_file, kd_file, ts_file, wt_file, mf_file, goldnet
def cross_timeseries_data(ds1, ds2):
""" Based on Piotr Mirovski's DFG4GRN code """
return