-
Notifications
You must be signed in to change notification settings - Fork 3
/
RMSE-Milnes-compare-plot.py
executable file
·185 lines (164 loc) · 8.91 KB
/
RMSE-Milnes-compare-plot.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
#!/usr/bin/env python3
import sys
sys.path.append('..')
import os
import argparse
import numpy as np
import matplotlib.pyplot as plt
plt.rcParams['axes.axisbelow'] = True
from methods import results
results = os.path.join(results, 'RMSE-Milnes')
if not os.path.isdir(results):
os.makedirs(results)
path2fits = '../data/Milnes-data-fits'
path2fitsre = '../data/Milnes-data-fits-re'
path2data = '../data/Raw-Milnes-data'
drugs = ['astemizole', 'azimilide', 'bepridil', 'chlorpromazine', 'cisapride', 'clarithromycin', 'clozapine', \
'diltiazem', 'disopyramide', 'dofetilide', 'domperidone', 'droperidol', 'ibutilide', 'loratadine', \
'metoprolol', 'mexiletine', 'nifedipine', 'nitrendipine', 'ondansetron', 'pimozide', 'quinidine', \
'ranolazine', 'risperidone', 'sotalol', 'tamoxifen', 'terfenadine', 'vandetanib', 'verapamil']
# Check input arguments
parser = argparse.ArgumentParser(
description='Plot model and experimental data')
parser.add_argument("--drug", type=str, choices=drugs, default='quinidine')
parser.add_argument("--all", action='store_true', help="Plot all drugs, this will ignore --drug argument", \
default=False)
parser.add_argument("--show", action='store_true', help="Show plots", \
default=False)
parser.add_argument("--n_samples", type=int, help="Number of Bootstrap samples to use", \
default=1000)
args = parser.parse_args()
baseline_models = ['lei', 'fda']
xticks_val = [r'0$\alpha$', r'0$\beta$', 1, 2, '2i', 3, 4, 5, '5i', 6, 7, 8, 9, 10, 11, 12, 13, '']
for i in range(1, len(xticks_val), 2):
xticks_val[i] = '\n' + str(xticks_val[i])
nmodels = len(xticks_val) - 1
RMSEs = np.zeros((len(baseline_models), nmodels))
models = np.linspace(1, nmodels + 1, nmodels + 1, dtype=int)
if args.all:
plot_drugs = drugs
else:
plot_drugs = [args.drug]
for drug in plot_drugs:
for i, m in enumerate(models[:13]):
model_str = 'model' + str(m)
for a, b in enumerate(baseline_models):
RMSE = np.loadtxt(path2fits + '/Drug-' + drug + '-' + b + '-' + model_str + \
'-protocol-Milnes-fit-RMSE-errors.txt')
if (RMSE.size > 1):
RMSE = RMSE[0]
RMSEs[a][i] = RMSE
# NOTE: Match xticks_val[:-1]
RMSEs[:, 9:] = RMSEs[:, 5:13]
RMSEs[:, 5:8] = RMSEs[:, 2:5]
RMSEs[:, 2:4] = RMSEs[:, :2]
for i, m in zip([0, 1, 4, 8], ['0a', '0b', '2i', '5i']):
for a, b in enumerate(baseline_models):
if b == 'lei':
RMSE = np.loadtxt(path2fitsre + f'/{drug}-m{m}-errors.txt')
elif b == 'fda':
RMSE = np.loadtxt(path2fitsre + f'/{drug}-li-m{m}-errors.txt')
else:
raise ValueError(f'Unexpected model {b}')
if (RMSE.size > 1):
RMSE = RMSE[0]
RMSEs[a][i] = RMSE
FDA_RMSE = np.loadtxt(path2fits + '/FDA/' + drug + '-RMSE.txt')
expt_samples = np.loadtxt(path2data + '/' + drug + '-bootstrap-' + str(args.n_samples) + '-samples.txt')
colors = ['#1f77b4', '#ff7f0e', '#2ca02c', '#d62728', '#9467bd', '#8c564b', '#e377c2', '#7f7f7f', '#bcbd22']
xs = np.random.normal(1.0, 0.05, len(expt_samples))
show_all_samples = False
fig, (a0, a1) = plt.subplots(1, 2, gridspec_kw={'width_ratios': [1, 4]}, figsize=(8.5, 4), sharey=True)
plt.title(drug)
B = a0.boxplot(expt_samples, sym='', whis=10)
Q0 = np.min([item.get_ydata()[1] for item in B['whiskers']])
Q4 = np.max([item.get_ydata()[1] for item in B['whiskers']])
if show_all_samples:
for d, e in enumerate(expt_samples):
if e >= Q0 and e <= Q4:
a0.scatter(xs[d], e, alpha=0.2, color='dodgerblue')
else:
a0.scatter(xs[d], e, alpha=0.2, color='red')
a0.scatter(xs, expt_samples, alpha=0.2, color='dodgerblue')
a0.axhline(Q4, color='silver', linestyle='dashed', zorder=-1)
a0.set_ylabel('RMSD')
a0.set_xticklabels(['Bootstrap\nsamples'])
a0.grid(True)
a1.scatter(models[0], RMSEs[1][0], marker='s', s=100, color='none', linewidth=1.5, \
edgecolor=colors[0], label='Physiological model A')
a1.scatter(models[0], RMSEs[0][0], marker='o', s=100, color='none', linewidth=1.5, \
edgecolor=colors[0], label='Physiological model B')
#a1.scatter(models[0], RMSEs[2][0], marker='d', s=100, color='none', linewidth=1.5, \
# edgecolor='k', label='Whittaker model')
a1.scatter(models[1], RMSEs[1][1], marker='s', s=100, color='none', linewidth=1.5, \
edgecolor='none', label='\n')
#a1.scatter(models[1], RMSEs[0][1], marker='o', s=100, color='none', linewidth=1.5, \
# edgecolor='k', label=r'2 params: $k_{\rm{on}}$, $k_{\rm{off}}$')
a1.scatter(models[1:9], RMSEs[0][1:9], marker='o', s=100, color='none', linewidth=1.5, \
edgecolor=colors[0], label=r'3 params: $k_{\rm{on}}$, $k_{\rm{off}}$, $n$')
a1.scatter(models[9], RMSEs[0][9], marker='o', s=100, color='none', linewidth=1.5, \
edgecolor=colors[1], label=r'4 params: $k_{\rm{on}}$, $k_{\rm{off}}$, $k_{\rm{trap}}$, $n$')
a1.scatter(models[10], RMSEs[0][10], marker='o', s=100, color='none', linewidth=1.5, \
edgecolor=colors[2], label=r'5 params: $k_{\rm{on, O}}$, $k_{\rm{off, O}}$, $k_{\rm{on, I}}$, $k_{\rm{off, I}}$, $n$')
a1.scatter(models[11:13], RMSEs[0][11:13], marker='o', s=100, color='none', linewidth=1.5, \
edgecolor=colors[3], label=r'4 params: $k_{\rm{on, O}}$, $k_{\rm{off, O}}$, $k_{\rm{on, I}}$, $n$')
a1.scatter(models[13], RMSEs[0][13], marker='o', s=100, color='none', linewidth=1.5, \
edgecolor=colors[4], label=r'5 params: $k_{\rm{on, O}}$, $k_{\rm{off, O}}$, $k_{\rm{on, I}}$, $k_{\rm{trap}}$, $n$')
a1.scatter(models[14], RMSEs[0][14], marker='o', s=100, color='none', linewidth=1.5, \
edgecolor=colors[5], label=r'5 params: $k_{\rm{on}}$, $k_{\rm{off}}$, $k_{\rm{untrap}}$, $k_{\rm{trap}}$, $n$')
a1.scatter(models[15], RMSEs[0][15], marker='o', s=100, color='none', linewidth=1.5, \
edgecolor=colors[6], label=r'5 params: $\hat{k}_{\rm{on}}$, $k_{\rm{off}}$, $\rm{EC}_{50}$, $n$, $V_{\rm{1/2, trap}}$')
a1.scatter(models[16], RMSEs[0][16], marker='o', s=100, color='none', linewidth=1.5, \
edgecolor=colors[7], label=r'4 params: $k_{\rm{on}}$, $k_{\rm{off}}$, $n$, $V_{\rm{1/2, trap}}$')
#a1.scatter(models[1], RMSEs[1][1], marker='s', s=100, color='none', linewidth=1.5, \
# edgecolor='k')
a1.scatter(models[1:9], RMSEs[1][1:9], marker='s', s=100, color='none', linewidth=1.5, \
edgecolor=colors[0])
a1.scatter(models[9], RMSEs[1][9], marker='s', s=100, color='none', linewidth=1.5, \
edgecolor=colors[1])
a1.scatter(models[10], RMSEs[1][10], marker='s', s=100, color='none', linewidth=1.5, \
edgecolor=colors[2])
a1.scatter(models[11:13], RMSEs[1][11:13], marker='s', s=100, color='none', linewidth=1.5, \
edgecolor=colors[3])
a1.scatter(models[13], RMSEs[1][13], marker='s', s=100, color='none', linewidth=1.5, \
edgecolor=colors[4])
a1.scatter(models[14], RMSEs[1][14], marker='s', s=100, color='none', linewidth=1.5, \
edgecolor=colors[5])
a1.scatter(models[15], RMSEs[1][15], marker='s', s=100, color='none', linewidth=1.5, \
edgecolor=colors[6])
a1.scatter(models[16], RMSEs[1][16], marker='s', s=100, color='none', linewidth=1.5, \
edgecolor=colors[7])
'''
#a1.scatter(models[1], RMSEs[2][1], marker='d', s=100, color='none', linewidth=1.5, \
# edgecolor='k')
a1.scatter(models[1:9], RMSEs[2][1:9], marker='d', s=100, color='none', linewidth=1.5, \
edgecolor=colors[0])
a1.scatter(models[9], RMSEs[2][9], marker='d', s=100, color='none', linewidth=1.5, \
edgecolor=colors[1])
a1.scatter(models[10], RMSEs[2][10], marker='d', s=100, color='none', linewidth=1.5, \
edgecolor=colors[2])
a1.scatter(models[11:13], RMSEs[2][11:13], marker='d', s=100, color='none', linewidth=1.5, \
edgecolor=colors[3])
a1.scatter(models[13], RMSEs[2][13], marker='d', s=100, color='none', linewidth=1.5, \
edgecolor=colors[4])
a1.scatter(models[14], RMSEs[2][14], marker='d', s=100, color='none', linewidth=1.5, \
edgecolor=colors[5])
a1.scatter(models[15], RMSEs[2][15], marker='d', s=100, color='none', linewidth=1.5, \
edgecolor=colors[6])
a1.scatter(models[16], RMSEs[2][16], marker='d', s=100, color='none', linewidth=1.5, \
edgecolor=colors[7])
'''
a1.scatter(models[nmodels], FDA_RMSE, marker='*', s=100, color='none', linewidth=1.5, \
edgecolor='magenta', label=r'5 params: CiPA v1.0 model')
a1.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0., fontsize=7.5)
a1.axhline(Q4, color='silver', linestyle='dashed', zorder=-1)
a1.set_xlabel('Models')
a1.set_xticks(models, xticks_val)
a1.grid(True)
plt.tight_layout()
if args.show:
plt.show()
else:
plt.savefig(os.path.join(results, 'RMSE-' + drug + '-all-models'),
dpi=300)
plt.close()