-
Notifications
You must be signed in to change notification settings - Fork 1
/
distances.py
355 lines (327 loc) · 16.1 KB
/
distances.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
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
"""Creates plots to visualize cell clustering data"""
import pandas as pd
import glob
import seaborn as sns
import numpy as np
import math
from astropy.stats import RipleysKEstimator
def PlotSingleDistances(folder, extension, ax, log=False):
"""Plots boxplot of distance to closest cell give ImageJ data, with or without log transformation"""
times = GetTimes(folder, extension)
file_frame = pd.concat(Generate_dfs(folder, extension, times))
if log:
logs = []
for length in file_frame["Length"]:
logs.append(math.log(length))
file_frame["Log Lengths"] = logs
sns.boxplot(x="Time", y="Log Lengths", data=file_frame, ax=ax)
ax.set_ylim(-1.7, 1.6)
ax.set_title(folder)
else:
sns.boxplot(x="Time", y="Length", data=file_frame, ax=ax)
ax.set_ylim(0, 4.5)
ax.set_title(folder)
def GetTimes(folder, extension):
"""Takes in a folder and extension in correct format and returns list of times"""
filenames = glob.glob("msresist/data/Distances/" + folder + "/Results_" + extension + "*.csv")
filename_prefix = "msresist/data/Distances/" + folder + "/Results_" + extension + "_"
filename_suffix = ".csv"
times = []
for file in filenames:
time = int(file[len(filename_prefix): -len(filename_suffix)])
times.append(time)
times = sorted(times)
return times
def Generate_dfs(folder, extension, times):
"""Generates dfs of the data at each time point with an added column for time"""
file_list = []
for time in times:
file = pd.read_csv("msresist/data/Distances/" + folder + "/Results_" + extension + "_" + str(time) + ".csv")
file["Time"] = time
file_list.append(file)
return file_list
def Calculate_closest(file_list, n=(1, 3)):
"""Calculates distances to nearby cells and returns as dataframe ready to plot"""
distances_by_time = []
for idx, file in enumerate(file_list):
distances_df = pd.DataFrame()
points = file.loc[:, "X":"Y"]
# shortest_n_distances_lists = []
shortest_n_distances = []
for origin_cell in np.arange(points.shape[0]):
distances = []
x1, y1 = points.iloc[origin_cell, :]
for other_cell in np.arange(points.shape[0]):
x2, y2 = points.iloc[other_cell, :]
distance = abs(math.sqrt((x2 - x1) ** 2 + (y2 - y1) ** 2))
distances.append(distance)
distances = sorted(distances)
# shortest_n_distances_lists.append(distances[1:(n+1)])
shortest_n_distances.extend(distances[n[0]: (n[1] + 1)])
distances_df["Distances"] = shortest_n_distances
distances_df["Time"] = idx * 3
distances_by_time.append(distances_df)
return pd.concat(distances_by_time)
def PlotClosestN(folder, extension, ax, log=False, cells=(1, 3)):
"""Plots specified range of nearby cells as boxplots with or without log transformation"""
times = GetTimes(folder, extension)
file_list = Generate_dfs(folder, extension, times)
plotting_frame = Calculate_closest(file_list, cells)
if log:
logs = []
for length in plotting_frame["Distances"]:
logs.append(math.log(length))
plotting_frame["Log Distances"] = logs
sns.boxplot(x="Time", y="Log Distances", data=plotting_frame, ax=ax)
ax.set_ylim(-2, 2.3)
ax.set_title(extension)
else:
sns.boxplot(x="Time", y="Distances", data=plotting_frame, ax=ax)
ax.set_ylim(0, 8)
ax.set_title(extension)
def PlotNhrsdistances(folder, mutants, treatments, replicates, ax, log=False, logmean=False, cells=(1, 3)):
"""Creates either raw/log grouped boxplot or log pointplot for distances to cells depending on log and logmean variables for range of nearby cells"""
dfs = []
for mutant in mutants:
mut_frames = []
for treatment in treatments:
for replicate in range(1, replicates + 1):
if replicate != 1:
file = pd.read_csv("msresist/data/Distances/" + folder + "/Results_" + mutant + treatment + str(replicate) + ".csv")
else:
file = pd.read_csv("msresist/data/Distances/" + folder + "/Results_" + mutant + treatment + ".csv")
distances = calculatedistances(file, mutant, treatment, replicate, cells)
mut_frames.append(distances)
mut_frame = pd.concat(mut_frames)
dfs.append(mut_frame)
to_plot = pd.concat(dfs)
if log:
logs = []
for length in to_plot["Distances"]:
logs.append(math.log(length))
to_plot["Log Distances"] = logs
if logmean:
sns.pointplot(x="Mutant", y="Log Distances", hue="Condition", data=to_plot, ci=68, join=False, dodge=0.4, ax=ax)
else:
sns.boxplot(x="Mutant", y="Log Distances", hue="Condition", data=to_plot, ax=ax)
else:
sns.boxplot(x="Mutant", y="Distances", hue="Condition", data=to_plot, ax=ax)
ax.set_ylim(0, 5)
# b = sns.swarmplot(x='Mutant', y='Distances', hue='Condition', data=to_plot, dodge=True, ax=ax)
def calculatedistances(file, mutant, treatment, replicate, cells=(1, 3)):
"""Calculates distances to range of other cells for a given mutant, treatment, and condition"""
distances_df = pd.DataFrame()
points = file.loc[:, "X":"Y"]
shortest_n_distances = []
for origin_cell in np.arange(points.shape[0]):
distances = []
x1, y1 = points.iloc[origin_cell, :]
for other_cell in np.arange(points.shape[0]):
x2, y2 = points.iloc[other_cell, :]
distance = abs(math.sqrt((x2 - x1) ** 2 + (y2 - y1) ** 2))
distances.append(distance)
distances = sorted(distances)
shortest_n_distances.extend(distances[cells[0]: (cells[1] + 1)])
distances_df["Distances"] = shortest_n_distances
distances_df["Mutant"] = mutant
if replicate != 1:
distances_df["Condition"] = treatment + str(replicate)
else:
distances_df["Condition"] = treatment
return distances_df
def Plot_Logmean(folder, mutants, treatments, replicates, ax, vs_count=False, cells=(1, 3)):
"""Plots the log mean distance to neighbors by mutant and condition as given by cells argument.
Will plot vs number of cells in image if vs_count is True"""
dfs = []
for mutant in mutants:
mut_frames = []
for treatment in treatments:
for replicate in range(1, replicates + 1):
if replicate != 1:
file = pd.read_csv("msresist/data/Distances/" + folder + "/Results_" + mutant + treatment + str(replicate) + ".csv")
else:
file = pd.read_csv("msresist/data/Distances/" + folder + "/Results_" + mutant + treatment + ".csv")
distances = calculatedistances_logmean(file, mutant, treatment, vs_count, cells)
mut_frames.append(distances)
mut_frame = pd.concat(mut_frames)
dfs.append(mut_frame)
to_plot = pd.concat(dfs)
if vs_count:
sns.scatterplot(x="Cells", y="Log_Mean_Distances", hue="Condition", style="Condition", data=to_plot, ax=ax)
# for line in range(0, to_plot.shape[0]):
# b.text(to_plot.Cells.iloc[line], to_plot.Log_Mean_Distances.iloc[line], to_plot.Mutant.iloc[line], horizontalalignment='left', size='medium', color='black', weight='semibold')
else:
sns.pointplot(x="Mutant", y="Log_Mean_Distances", hue="Condition", data=to_plot, ci=68, join=False, dodge=0.25, ax=ax)
def calculatedistances_logmean(file, mutant, treatment, vs_count, cells=(1, 3)):
"""Calculates the average log distance to neighbors as defined by the cells argument and returns as a DataFrame in proper plotting format"""
points = file.loc[:, "X":"Y"]
shortest_n_distances = []
for origin_cell in np.arange(points.shape[0]):
distances = []
x1, y1 = points.iloc[origin_cell, :]
for other_cell in np.arange(points.shape[0]):
x2, y2 = points.iloc[other_cell, :]
distance = abs(math.sqrt((x2 - x1) ** 2 + (y2 - y1) ** 2))
distances.append(distance)
distances = sorted(distances)
shortest_n_distances.extend(distances[cells[0]: (cells[1] + 1)])
logs = []
for length in shortest_n_distances:
logs.append(math.log(length))
if vs_count:
distances_df = {"Log_Mean_Distances": [np.mean(logs)], "Cells": [points.shape[0]], "Condition": [treatment], "Mutant": [mutant]}
else:
distances_df = {"Log_Mean_Distances": [np.mean(logs)], "Mutant": [mutant], "Condition": [treatment]}
distances_df = pd.DataFrame(distances_df)
return distances_df
def PlotRipleysK(folder, mutant, treatments, replicates, ax):
"""Plots the Ripley's K Estimate in comparison to the Poisson for a range of radii"""
Kest = RipleysKEstimator(area=158.8761, x_max=14.67, y_max=10.83, x_min=0, y_min=0)
r = np.linspace(0, 5, 51)
r_for_df = r
poisson = Kest.poisson(r)
poisson_for_df = poisson
for i in range(5):
poisson_for_df = np.hstack((poisson_for_df, poisson))
r_for_df = np.hstack((r_for_df, r))
data = np.vstack((r_for_df, poisson_for_df))
for treatment in treatments:
reps = []
for replicate in range(1, replicates + 1):
if replicate != 1:
file = pd.read_csv("msresist/data/Distances/" + folder + "/Results_" + mutant + treatment + str(replicate) + ".csv")
else:
file = pd.read_csv("msresist/data/Distances/" + folder + "/Results_" + mutant + treatment + ".csv")
points = file.loc[:, "X":"Y"].values
reps.append(points)
Kests = []
for point_set in reps:
Kests.append(Kest(data=point_set, radii=r, mode="ripley"))
treat_array = np.hstack((Kests[0], Kests[1]))
for i in range(2, len(Kests)):
treat_array = np.hstack((treat_array, Kests[i]))
data = np.vstack((data, treat_array))
df = pd.DataFrame(data).T
df.columns = ["Radii", "Poisson", "Untreated", "Erlotinib", "AF154 + Erlotinib"]
df = pd.melt(df, ["Radii"])
df.columns = ["Radii", "Condition", "K Estimate"]
sns.lineplot(x="Radii", y="K Estimate", hue="Condition", data=df, ci=68, ax=ax)
ax.set_title(mutant)
def BarPlotRipleysK(folder, mutants, xticklabels, treatments, legendlabels, replicates, r):
"""Plots a bar graph of the Ripley's K Estimate values for all mutants and conditions in comparison to the Poisson at a discrete radius.
Note that radius needs to be input as a 1D array for the RipleysKEstimator to work"""
Kest = RipleysKEstimator(area=158.8761, x_max=14.67, y_max=10.83, x_min=0, y_min=0)
poisson = Kest.poisson(r)
mutant_dfs = []
for z, mutant in enumerate(mutants):
for j, treatment in enumerate(treatments):
reps = []
for replicate in range(1, replicates + 1):
if replicate != 1:
file = pd.read_csv("msresist/data/Distances/" + folder + "/Results_" + mutant + treatment + str(replicate) + ".csv")
else:
file = pd.read_csv("msresist/data/Distances/" + folder + "/Results_" + mutant + treatment + ".csv")
points = file.loc[:, "X":"Y"].values
reps.append(points)
Kests = []
for point_set in reps:
Kests.append(Kest(data=point_set, radii=r, mode="ripley") / poisson)
if len(Kests) > 1:
treat_array = np.hstack((Kests[0], Kests[1]))
for i in range(2, len(Kests)):
treat_array = np.hstack((treat_array, Kests[i]))
else:
treat_array = Kests[0]
df = pd.DataFrame(treat_array)
df.columns = ["K Estimate"]
df["AXL mutants Y->F"] = xticklabels[z]
df["Treatment"] = legendlabels[j]
# poisson_df = pd.DataFrame(poisson)
# poisson_df.columns = ['K Estimate']
# poisson_df['Mutant'] = mutant
# poisson_df['Treatment'] = 'Poisson'
# df = pd.concat([poisson_df, df])
mutant_dfs.append(df)
df = pd.concat(mutant_dfs)
b = sns.barplot(x="AXL mutants Y->F", y="K Estimate", hue="Treatment", data=df, ci=68)
b.set_title("Radius of " + str(r[0]) + " Normalized to Poisson")
b.legend(loc=2)
def BarPlotRipleysK_TimePlots(folder, mutant, extensions, treatments, r, ax):
Kest = RipleysKEstimator(area=158.8761, x_max=14.67, y_max=10.83, x_min=0, y_min=0)
poisson = Kest.poisson(r)
treatment_dfs = []
for idx, extension in enumerate(extensions):
reps = []
file = pd.read_csv("msresist/data/Distances/" + folder + "/Results_" + extension + ".csv")
points = file.loc[:, "X":"Y"].values
treat_array = Kest(data=points, radii=r, mode="ripley") / poisson
df = pd.DataFrame(treat_array)
df.columns = ["K Estimate"]
df["Mutant"] = mutant
df["Treatment"] = treatments[idx]
# poisson_df = pd.DataFrame(poisson)
# poisson_df.columns = ['K Estimate']
# poisson_df['Mutant'] = mutant
# poisson_df['Treatment'] = 'Poisson'
# df = pd.concat([poisson_df, df])
treatment_dfs.append(df)
df = pd.concat(treatment_dfs)
sns.barplot(x="Mutant", y="K Estimate", hue="Treatment", data=df, ci=68, ax=ax)
def DataFrameRipleysK(folder, mutants, treatments, replicates, r):
Kest = RipleysKEstimator(area=158.8761, x_max=14.67, y_max=10.83, x_min=0, y_min=0)
poisson = Kest.poisson(r)
mutant_dfs = []
for mutant in mutants:
for treatment in treatments:
reps = []
for replicate in range(1, replicates + 1):
if replicate != 1:
file = pd.read_csv("msresist/data/Distances/" + folder + "/Results_" + mutant + treatment + str(replicate) + ".csv")
else:
file = pd.read_csv("msresist/data/Distances/" + folder + "/Results_" + mutant + treatment + ".csv")
points = file.loc[:, "X":"Y"].values
reps.append(points)
Kests = []
for point_set in reps:
Kests.append(Kest(data=point_set, radii=r, mode="ripley") / poisson)
if len(Kests) > 1:
treat_array = np.hstack((Kests[0], Kests[1]))
for i in range(2, len(Kests)):
treat_array = np.hstack((treat_array, Kests[i]))
else:
treat_array = Kests[0]
df = pd.DataFrame(treat_array)
df.columns = ["K Estimate"]
df["Mutant"] = mutant
df["Treatment"] = treatment
# poisson_df = pd.DataFrame(poisson)
# poisson_df.columns = ['K Estimate']
# poisson_df['Mutant'] = mutant
# poisson_df['Treatment'] = 'Poisson'
# df = pd.concat([poisson_df, df])
mutant_dfs.append(df)
df = pd.concat(mutant_dfs)
return df.groupby(["Mutant", "Treatment"]).mean()
def PlotRipleysK_TimeCourse(folder, extensions, timepoint, ax):
"""Plots the Ripley's K Estimate for a series of images over time by condition, compared to the Poisson."""
r = np.linspace(0, 5, 51)
Kest = RipleysKEstimator(area=158.8761, x_max=14.67, y_max=10.83, x_min=0, y_min=0)
poisson = Kest.poisson(r)
data = np.vstack((r, poisson))
treatments = []
for extension in extensions:
file = pd.read_csv("msresist/data/Distances/" + folder + "/Results_" + extension + "_" + str(timepoint) + ".csv")
points = file.loc[:, "X":"Y"].values
treatments.append(points)
Kests = []
for point_set in treatments:
Kests.append(Kest(data=point_set, radii=r, mode="ripley"))
treat_array = np.vstack((Kests[0], Kests[1]))
treat_array = np.vstack((treat_array, Kests[2]))
data = np.vstack((data, treat_array))
df = pd.DataFrame(data).T
df.columns = ["Radii", "Poisson", "Untreated", "Erlotinib", "Erlotinib + AF154"]
df = pd.melt(df, ["Radii"])
df.columns = ["Radii", "Condition", "K Estimate"]
sns.lineplot(x="Radii", y="K Estimate", hue="Condition", data=df, ci=68, ax=ax)
ax.set_title(str(timepoint) + " hours")