-
Notifications
You must be signed in to change notification settings - Fork 15
/
Copy path06R-Perform-machine-learning-on-Weizmann-data.R
executable file
·324 lines (289 loc) · 17.9 KB
/
06R-Perform-machine-learning-on-Weizmann-data.R
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
#-----------------------------------------------------------------------------
# 06-Perform-machine-learning-on-Weizmann-data.R
# Copyright (c) 2021--, Greg Poore
# Purposes:
# - Perform machine learning on Weizmann data
#-----------------------------------------------------------------------------
#----------------------------------------------------------#
# Load environments
#----------------------------------------------------------#
require(doMC)
require(plyr)
require(dplyr)
require(phyloseq)
require(microbiome)
require(vegan)
require(tibble)
require(biomformat)
require(rhdf5)
require(ggpubr)
require(ggsci)
require(scales)
numCores <- detectCores()
registerDoMC(cores=numCores)
#----------------------------------------------------------#
# Import data
#----------------------------------------------------------#
load("Interim_data/summarized_weizmann_data_various_taxa_levels_29Mar22.RData",
verbose = TRUE)
#----------------------------------------------------------#
# Convert data to relative abundance and binary formats
# and add a taxa diversity column
#----------------------------------------------------------#
##----------------------All features----------------------##
# Relative abundances
ra_weizmannPhylum <- weizmannPhylum/rowSums(weizmannPhylum)
ra_weizmannClass <- weizmannClass/rowSums(weizmannClass)
ra_weizmannOrder <- weizmannOrder/rowSums(weizmannOrder)
ra_weizmannFamily <- weizmannFamily/rowSums(weizmannFamily)
ra_weizmannGenus <- weizmannGenus/rowSums(weizmannGenus)
ra_weizmannSpecies <- weizmannSpecies/rowSums(weizmannSpecies)
# Binary formats (based on presence/absence)
binary_weizmannPhylum <- data.frame((weizmannPhylum>0)*1)
binary_weizmannClass <- data.frame((weizmannClass>0)*1)
binary_weizmannOrder <- data.frame((weizmannOrder>0)*1)
binary_weizmannFamily <- data.frame((weizmannFamily>0)*1)
binary_weizmannGenus <- data.frame((weizmannGenus>0)*1)
binary_weizmannSpecies <- data.frame((weizmannSpecies>0)*1)
# Add diversity columns
ra_weizmannPhylum$diversity <- rowSums(ra_weizmannPhylum>0)
ra_weizmannClass$diversity <- rowSums(ra_weizmannClass>0)
ra_weizmannOrder$diversity <- rowSums(ra_weizmannOrder>0)
ra_weizmannFamily$diversity <- rowSums(ra_weizmannFamily>0)
ra_weizmannGenus$diversity <- rowSums(ra_weizmannGenus>0)
ra_weizmannSpecies$diversity <- rowSums(ra_weizmannSpecies>0)
binary_weizmannPhylum$diversity <- rowSums(binary_weizmannPhylum)
binary_weizmannClass$diversity <- rowSums(binary_weizmannClass)
binary_weizmannOrder$diversity <- rowSums(binary_weizmannOrder)
binary_weizmannFamily$diversity <- rowSums(binary_weizmannFamily)
binary_weizmannGenus$diversity <- rowSums(binary_weizmannGenus)
binary_weizmannSpecies$diversity <- rowSums(binary_weizmannSpecies)
##----------------------Shared features----------------------##
# Relative abundances
ra_weizmannPhylumShared_Nonzero <- weizmannPhylumShared_Nonzero/rowSums(weizmannPhylumShared_Nonzero)
ra_weizmannClassShared_Nonzero <- weizmannClassShared_Nonzero/rowSums(weizmannClassShared_Nonzero)
ra_weizmannOrderShared_Nonzero <- weizmannOrderShared_Nonzero/rowSums(weizmannOrderShared_Nonzero)
ra_weizmannFamilyShared_Nonzero <- weizmannFamilyShared_Nonzero/rowSums(weizmannFamilyShared_Nonzero)
ra_weizmannGenusShared_Nonzero <- weizmannGenusShared_Nonzero/rowSums(weizmannGenusShared_Nonzero)
ra_weizmannSpeciesShared_Nonzero <- weizmannSpeciesShared_Nonzero/rowSums(weizmannSpeciesShared_Nonzero)
# Binary formats (based on presence/absence)
binary_weizmannPhylumShared_Nonzero <- data.frame((weizmannPhylumShared_Nonzero>0)*1)
binary_weizmannClassShared_Nonzero <- data.frame((weizmannClassShared_Nonzero>0)*1)
binary_weizmannOrderShared_Nonzero <- data.frame((weizmannOrderShared_Nonzero>0)*1)
binary_weizmannFamilyShared_Nonzero <- data.frame((weizmannFamilyShared_Nonzero>0)*1)
binary_weizmannGenusShared_Nonzero <- data.frame((weizmannGenusShared_Nonzero>0)*1)
binary_weizmannSpeciesShared_Nonzero <- data.frame((weizmannSpeciesShared_Nonzero>0)*1)
# Add diversity columns
ra_weizmannPhylumShared_Nonzero$diversity <- rowSums(ra_weizmannPhylumShared_Nonzero>0)
ra_weizmannClassShared_Nonzero$diversity <- rowSums(ra_weizmannClassShared_Nonzero>0)
ra_weizmannOrderShared_Nonzero$diversity <- rowSums(ra_weizmannOrderShared_Nonzero>0)
ra_weizmannFamilyShared_Nonzero$diversity <- rowSums(ra_weizmannFamilyShared_Nonzero>0)
ra_weizmannGenusShared_Nonzero$diversity <- rowSums(ra_weizmannGenusShared_Nonzero>0)
ra_weizmannSpeciesShared_Nonzero$diversity <- rowSums(ra_weizmannSpeciesShared_Nonzero>0)
binary_weizmannPhylumShared_Nonzero$diversity <- rowSums(binary_weizmannPhylumShared_Nonzero)
binary_weizmannClassShared_Nonzero$diversity <- rowSums(binary_weizmannClassShared_Nonzero)
binary_weizmannOrderShared_Nonzero$diversity <- rowSums(binary_weizmannOrderShared_Nonzero)
binary_weizmannFamilyShared_Nonzero$diversity <- rowSums(binary_weizmannFamilyShared_Nonzero)
binary_weizmannGenusShared_Nonzero$diversity <- rowSums(binary_weizmannGenusShared_Nonzero)
binary_weizmannSpeciesShared_Nonzero$diversity <- rowSums(binary_weizmannSpeciesShared_Nonzero)
##----------------------Add diversity columns to count data----------------------##
## Add diversity columns -- CANNOT add before the above code (makes the the RA calculations incorrect)
weizmannPhylum$diversity <- rowSums(weizmannPhylum>0)
weizmannClass$diversity <- rowSums(weizmannClass>0)
weizmannOrder$diversity <- rowSums(weizmannOrder>0)
weizmannFamily$diversity <- rowSums(weizmannFamily>0)
weizmannGenus$diversity <- rowSums(weizmannGenus>0)
weizmannSpecies$diversity <- rowSums(weizmannSpecies>0)
weizmannPhylumShared_Nonzero$diversity <- rowSums(weizmannPhylumShared_Nonzero>0)
weizmannClassShared_Nonzero$diversity <- rowSums(weizmannClassShared_Nonzero>0)
weizmannOrderShared_Nonzero$diversity <- rowSums(weizmannOrderShared_Nonzero>0)
weizmannFamilyShared_Nonzero$diversity <- rowSums(weizmannFamilyShared_Nonzero>0)
weizmannGenusShared_Nonzero$diversity <- rowSums(weizmannGenusShared_Nonzero>0)
weizmannSpeciesShared_Nonzero$diversity <- rowSums(weizmannSpeciesShared_Nonzero>0)
save(weizmannPhylum,
ra_weizmannPhylum,
binary_weizmannPhylum,
weizmannMetaPhylum,
weizmannClass,
ra_weizmannClass,
binary_weizmannClass,
weizmannMetaClass,
weizmannOrder,
ra_weizmannOrder,
binary_weizmannOrder,
weizmannMetaOrder,
weizmannFamily,
ra_weizmannFamily,
binary_weizmannFamily,
weizmannMetaFamily,
weizmannGenus,
ra_weizmannGenus,
binary_weizmannGenus,
weizmannMetaGenus,
weizmannSpecies,
ra_weizmannSpecies,
binary_weizmannSpecies,
weizmannMetaSpecies,
weizmannPhylumShared_Nonzero,
ra_weizmannPhylumShared_Nonzero,
binary_weizmannPhylumShared_Nonzero,
weizmannMetaPhylumShared_Nonzero,
weizmannClassShared_Nonzero,
ra_weizmannClassShared_Nonzero,
binary_weizmannClassShared_Nonzero,
weizmannMetaClassShared_Nonzero,
weizmannOrderShared_Nonzero,
ra_weizmannOrderShared_Nonzero,
binary_weizmannOrderShared_Nonzero,
weizmannMetaOrderShared_Nonzero,
weizmannFamilyShared_Nonzero,
ra_weizmannFamilyShared_Nonzero,
binary_weizmannFamilyShared_Nonzero,
weizmannMetaFamilyShared_Nonzero,
weizmannGenusShared_Nonzero,
ra_weizmannGenusShared_Nonzero,
binary_weizmannGenusShared_Nonzero,
weizmannMetaGenusShared_Nonzero,
weizmannSpeciesShared_Nonzero,
ra_weizmannSpeciesShared_Nonzero,
binary_weizmannSpeciesShared_Nonzero,
weizmannMetaSpeciesShared_Nonzero,
file = "Interim_data/data_for_ml_weizmann_2Apr22.RData")
#----------------------------------------------------------#
# Run ML on Weizmann data
#----------------------------------------------------------#
# Use "Interim_data/data_for_ml_weizmann_2Apr22" to run
# "Supporting_scripts/S16R-ML-fungi-10k-rep1-weizmann.R"
# Which was run on a Slurm compute cluster using the submission script
# "Supporting_scripts/S16R-Submit-job.sh"
# Then download the final CSV file
# which has been placed under the "Interim_data" folder
# source('test-S11-ML-fungi-10k-rep1-weizmann.R')
#----------------------------------------------------------#
# Plot ML performance
#----------------------------------------------------------#
wzMLres <- read.csv("Interim_data/rep_perfFungi_10k_rep1_tcga_weizmann_with_and_without_shared_feat_ALL_2Apr22.csv")
wzMLres <- wzMLres[,!(colnames(wzMLres) == "X")]
colnames(wzMLres)[1:2] <- c("AUROC","AUPR")
# Phylum
wzMLres$datasetName[wzMLres$datasetName == "ra_weizmannPhylum"] <- "Phylum (RA)"
wzMLres$datasetName[wzMLres$datasetName == "binary_weizmannPhylum"] <- "Phylum (Binary)"
wzMLres$datasetName[wzMLres$datasetName == "weizmannPhylum"] <- "Phylum (Raw Counts)"
wzMLres$datasetName[wzMLres$datasetName == "ra_weizmannPhylumShared_Nonzero"] <- "Phylum Shared (RA)"
wzMLres$datasetName[wzMLres$datasetName == "binary_weizmannPhylumShared_Nonzero"] <- "Phylum Shared (Binary)"
wzMLres$datasetName[wzMLres$datasetName == "weizmannPhylumShared_Nonzero"] <- "Phylum Shared (Raw Counts)"
# Class
wzMLres$datasetName[wzMLres$datasetName == "ra_weizmannClass"] <- "Class (RA)"
wzMLres$datasetName[wzMLres$datasetName == "binary_weizmannClass"] <- "Class (Binary)"
wzMLres$datasetName[wzMLres$datasetName == "weizmannClass"] <- "Class (Raw Counts)"
wzMLres$datasetName[wzMLres$datasetName == "ra_weizmannClassShared_Nonzero"] <- "Class Shared (RA)"
wzMLres$datasetName[wzMLres$datasetName == "binary_weizmannClassShared_Nonzero"] <- "Class Shared (Binary)"
wzMLres$datasetName[wzMLres$datasetName == "weizmannClassShared_Nonzero"] <- "Class Shared (Raw Counts)"
# Order
wzMLres$datasetName[wzMLres$datasetName == "ra_weizmannOrder"] <- "Order (RA)"
wzMLres$datasetName[wzMLres$datasetName == "binary_weizmannOrder"] <- "Order (Binary)"
wzMLres$datasetName[wzMLres$datasetName == "weizmannOrder"] <- "Order (Raw Counts)"
wzMLres$datasetName[wzMLres$datasetName == "ra_weizmannOrderShared_Nonzero"] <- "Order Shared (RA)"
wzMLres$datasetName[wzMLres$datasetName == "binary_weizmannOrderShared_Nonzero"] <- "Order Shared (Binary)"
wzMLres$datasetName[wzMLres$datasetName == "weizmannOrderShared_Nonzero"] <- "Order Shared (Raw Counts)"
# Family
wzMLres$datasetName[wzMLres$datasetName == "ra_weizmannFamily"] <- "Family (RA)"
wzMLres$datasetName[wzMLres$datasetName == "binary_weizmannFamily"] <- "Family (Binary)"
wzMLres$datasetName[wzMLres$datasetName == "weizmannFamily"] <- "Family (Raw Counts)"
wzMLres$datasetName[wzMLres$datasetName == "ra_weizmannFamilyShared_Nonzero"] <- "Family Shared (RA)"
wzMLres$datasetName[wzMLres$datasetName == "binary_weizmannFamilyShared_Nonzero"] <- "Family Shared (Binary)"
wzMLres$datasetName[wzMLres$datasetName == "weizmannFamilyShared_Nonzero"] <- "Family Shared (Raw Counts)"
# Genus
wzMLres$datasetName[wzMLres$datasetName == "ra_weizmannGenus"] <- "Genus (RA)"
wzMLres$datasetName[wzMLres$datasetName == "binary_weizmannGenus"] <- "Genus (Binary)"
wzMLres$datasetName[wzMLres$datasetName == "weizmannGenus"] <- "Genus (Raw Counts)"
wzMLres$datasetName[wzMLres$datasetName == "ra_weizmannGenusShared_Nonzero"] <- "Genus Shared (RA)"
wzMLres$datasetName[wzMLres$datasetName == "binary_weizmannGenusShared_Nonzero"] <- "Genus Shared (Binary)"
wzMLres$datasetName[wzMLres$datasetName == "weizmannGenusShared_Nonzero"] <- "Genus Shared (Raw Counts)"
# Species
wzMLres$datasetName[wzMLres$datasetName == "ra_weizmannSpecies"] <- "Species (RA)"
wzMLres$datasetName[wzMLres$datasetName == "binary_weizmannSpecies"] <- "Species (Binary)"
wzMLres$datasetName[wzMLres$datasetName == "weizmannSpecies"] <- "Species (Raw Counts)"
wzMLres$datasetName[wzMLres$datasetName == "ra_weizmannSpeciesShared_Nonzero"] <- "Species Shared (RA)"
wzMLres$datasetName[wzMLres$datasetName == "binary_weizmannSpeciesShared_Nonzero"] <- "Species Shared (Binary)"
wzMLres$datasetName[wzMLres$datasetName == "weizmannSpeciesShared_Nonzero"] <- "Species Shared (Raw Counts)"
wzMLres$datasetName <- factor(wzMLres$datasetName,
levels = c("Phylum (Raw Counts)", "Phylum (RA)", "Phylum (Binary)",
"Phylum Shared (Raw Counts)", "Phylum Shared (RA)", "Phylum Shared (Binary)",
"Class (Raw Counts)", "Class (RA)", "Class (Binary)",
"Class Shared (Raw Counts)", "Class Shared (RA)", "Class Shared (Binary)",
"Order (Raw Counts)", "Order (RA)", "Order (Binary)",
"Order Shared (Raw Counts)", "Order Shared (RA)", "Order Shared (Binary)",
"Family (Raw Counts)", "Family (RA)", "Family (Binary)",
"Family Shared (Raw Counts)", "Family Shared (RA)", "Family Shared (Binary)",
"Genus (Raw Counts)", "Genus (RA)", "Genus (Binary)",
"Genus Shared (Raw Counts)", "Genus Shared (RA)", "Genus Shared (Binary)",
"Species (Raw Counts)", "Species (RA)", "Species (Binary)",
"Species Shared (Raw Counts)", "Species Shared (RA)", "Species Shared (Binary)"))
#-------------------------Plot primary tumor 1 vs. all others performance-------------------------#
# There are lots of options to play with here. For purposes of the paper, we are focusing on the
# best performing option: full binary (presence/absence) data at the species level
source('Supporting_scripts/S00-SummarySE.R')
wzMLres %>%
filter(sampleType == "tumor") %>%
filter(!grepl("Shared",datasetName)) %>%
filter(grepl("Binary",datasetName)) %>%
filter(grepl("Species",datasetName)) %>%
droplevels() %>%
reshape2::melt(id.vars = c("rep","diseaseType","sampleType","datasetName","metadataName","minorityClassSize","majorityClassSize","minorityClassName","majorityClassName")) %>%
summarySE(measurevar = "value", groupvars = c("diseaseType","datasetName","variable")) %>%
ggplot(aes(reorder(diseaseType, value, FUN=median),value, color=variable)) +
geom_errorbar(aes(ymin=value-ci, ymax=value+ci), width=0, position = position_dodge(0.9)) +
geom_point(position = position_dodge(0.9), size=2.5) + xlab("Cancer type") + ylab("Area Under Curve") + theme_pubr() +
# facet_wrap(~variable) +
scale_y_continuous(breaks = seq(0, 1, by = 0.1), limits = c(0,1)) +
ggtitle("Tumor | 1 Vs All | Binary (Presence/Absence) data | Species Level") + theme(plot.title = element_text(hjust = 0.5)) +
# rotate_x_text(90) +
scale_color_nejm(name = "Features") +
geom_hline(yintercept = 1, linetype="dashed")
ggsave("Figures/Other_Figures/ml_weizmann_tumor_1vsAll_species_binary.pdf", dpi = "retina",
width = 10, height = 6, units = "in")
# There are lots of options to play with here. For purposes of the paper, we are focusing on the
# best performing option: full binary (presence/absence) data at the species level
source('Supporting_scripts/S00-SummarySE.R')
wzMLres %>%
filter(sampleType == "tumor") %>%
filter(!grepl("Shared",datasetName)) %>%
# filter(grepl("Binary",datasetName)) %>%
filter(grepl("Species",datasetName)) %>%
droplevels() %>%
reshape2::melt(id.vars = c("rep","diseaseType","sampleType","datasetName","metadataName","minorityClassSize","majorityClassSize","minorityClassName","majorityClassName")) %>%
summarySE(measurevar = "value", groupvars = c("diseaseType","datasetName","variable")) %>%
ggplot(aes(reorder(diseaseType, value, FUN=median),value, color=datasetName)) +
geom_errorbar(aes(ymin=value-ci, ymax=value+ci), width=0, position = position_dodge(0.9)) +
geom_point(position = position_dodge(0.9), size=2.5) + xlab("Cancer type") + ylab("Area Under Curve") + theme_pubr() +
facet_wrap(~variable) +
scale_y_continuous(breaks = seq(0, 1, by = 0.1), limits = c(0,1)) +
ggtitle("Tumor | 1 Vs All | Binary (Presence/Absence) data | Species Level") + theme(plot.title = element_text(hjust = 0.5)) +
# rotate_x_text(90) +
scale_color_nejm(name = "Features") +
geom_hline(yintercept = 1, linetype="dashed")
ggsave("Figures/Other_Figures/ml_weizmann_tumor_1vsAll_species_counts_ra_binary.pdf", dpi = "retina",
width = 10, height = 6, units = "in")
#-------------------------Plot primary tumor vs. nat performance-------------------------#
# No options appeared to differentiate tumor vs. nat well here, so all are shown
source('Supporting_scripts/S00-SummarySE.R')
wzMLres %>%
filter(sampleType == "tumor vs nat") %>%
filter(!grepl("Shared",datasetName)) %>%
# filter(grepl("Binary",datasetName)) %>%
# filter(grepl("Species",datasetName)) %>%
droplevels() %>%
reshape2::melt(id.vars = c("rep","diseaseType","sampleType","datasetName","metadataName","minorityClassSize","majorityClassSize","minorityClassName","majorityClassName")) %>%
summarySE(measurevar = "value", groupvars = c("diseaseType","datasetName","variable")) %>%
ggplot(aes(reorder(diseaseType, value, FUN=median),value, color=datasetName)) +
geom_errorbar(aes(ymin=value-ci, ymax=value+ci), width=0, position = position_dodge(0.9)) +
geom_point(position = position_dodge(0.9), size=2.5) + xlab("Cancer type") + ylab("Area Under Curve") + theme_pubr() +
facet_wrap(~variable) +
scale_y_continuous(breaks = seq(0, 1, by = 0.1), limits = c(0,1)) +
ggtitle("Tumor vs. NAT | Phylum, Class, Order, Family, Genus, Species Levels") + theme(plot.title = element_text(hjust = 0.5)) +
# rotate_x_text(90) +
scale_color_igv(name = "Features") +
geom_hline(yintercept = 1, linetype="dashed")
ggsave("Figures/Other_Figures/ml_weizmann_tumor_vs_nat_all.pdf", dpi = "retina",
width = 10, height = 6, units = "in")