-
Notifications
You must be signed in to change notification settings - Fork 0
/
plotting.R
225 lines (184 loc) · 7.81 KB
/
plotting.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
library(ggplot2)
library(grid)
library(gtools)
library(gridExtra)
# Solarized Color Pallete
solarized <- c(
yellow='#b58900', orange='#cb4b16', red='#dc322f', magenta='#d33682',
violet='#6c71c4', blue='#268bd2', cyan='#2aa198', green='#859900',
base03='#002b36', base02='#073642', base01='#586e75', base00='#657b83',
base0='#839496', base1='#93a1a1', base2='#eee8d5', base3='#fdf6e3')
Solar <- function(...) {
color.names <- c(...)
hex.cols <- as.character(solarized[color.names])
return(hex.cols)
}
# plotting parameters
width.full <- 6.83
width.half <- 3.27
breaks.roc <- seq(0, 1, by=0.2)
breaks.prc <- seq(0, 1, by=0.2)
ylim.prc <- c(0, 1)
# Colors
strip.fill <- Solar('base3')
'#300A24' # Blackberry (ubuntu terminal)
'#F8F6F7' # lighter blackberry
## Plotting functions
OpenPDF <- function(path, width=4, height=4) {
cairo_pdf(path, width=width, height=height, bg='white', family='sans')
}
ClosePDF <- function(path) {
dev.off()
embedFonts(path)
}
ChrRound <- function(x, digits=2) {
# Round x to digits and return as a character with trailing zeros.
x.rounded <- round(x, digits)
sprintf.chr <- sprintf('%%.%sf', digits)
x.chr <- sprintf(sprintf.chr, x.rounded)
return(x.chr)
}
SetGGTheme <- function(gg) {
gg <- gg + theme_bw() + theme(legend.title=element_text(face='plain'))
gg <- gg + theme(plot.margin=grid::unit(c(2, 2, 2, 2), 'points'))
gg <- gg + theme(legend.margin=grid::unit(0, 'cm'))
gg <- gg + theme(legend.key.height=grid::unit(0.90, 'lines'))
gg <- gg + theme(strip.background=element_rect(fill=strip.fill))
return(gg)
}
ggROC <- function(gg) {
gg <- SetGGTheme(gg) +
geom_segment(x=0, xend=1, y=0, yend=1, color='grey', size=0.15, show_guide=FALSE) +
theme(legend.background=element_rect(color='grey60', size=0.2)) +
theme(legend.key.width=grid::unit(1.5, 'lines')) +
theme(legend.key=element_rect(linetype='blank')) +
theme(legend.justification=c(1, 0), legend.position=c(1, 0)) +
scale_x_continuous(breaks=breaks.roc, expand=c(0.03, 0)) +
scale_y_continuous(breaks=breaks.roc, expand=c(0.03, 0)) +
xlab('False Positive Rate') + ylab('Recall') + coord_fixed()
return(gg)
}
ggPRC <- function(gg) {
gg <- SetGGTheme(gg) +
geom_path(size=.8, color='grey') + geom_point(size=1.5) +
theme(legend.justification=c(1, 1), legend.position=c(1, 1)) +
theme(legend.background=element_rect(color='grey60', size=0.2)) +
xlab('Recall') + ylab('Precision') +
scale_x_continuous(breaks=breaks.prc, expand=c(0.03, 0)) +
scale_y_continuous(breaks=breaks.prc, expand=c(0.03, 0), limits=ylim.prc) +
#scale_color_gradientn(colours = rainbow(7)[6:1]) +
scale_color_gradientn(colours = as.character(solarized[8:2])) +
guides(color=guide_colorbar(title='Prediction\nThreshold', nbin=500))
return(gg)
}
FacetWrapLabeller <- function(gg.plot, label.list=NULL) {
#works with R 3.0.1 and ggplot2 0.9.3.1
require(gridExtra)
g <- ggplotGrob(gg.plot)
gg <- g$grobs
strips <- grep('strip_t', names(gg))
for(ii in seq_along(label.list)) {
modgrob <- grid::getGrob(gg[[strips[ii]]], 'strip.text',
grep=TRUE, global=TRUE)
gg[[strips[ii]]]$children[[modgrob$name]] <- grid::editGrob(modgrob, label=label.list[[ii]])
}
g$grobs <- gg
class(g) <- c('arrange', 'ggplot', class(g))
return(g)
}
pathophys.colors <- c('#005200', '#B20000', '#8F008F', '#0000B2', '#E68A00', 'black')
msig.strip.text <- 'Gene\u2014{MSigDB Collection}\u2014Gene\u2014Disease DWPC'
# AUROC plot helpers
PerfPlot <- function(gg, ymin.auroc) {
gg <- SetGGTheme(gg) +
facet_grid(. ~ panel, scales='free_x', space='free_x') +
geom_hline(aes(yintercept=0.5), color='grey', linetype='solid') +
stat_summary(fun.y='MeanConfInt', geom='line', color='grey', size=11) +
theme(axis.text.x=element_text(angle=45, hjust=1)) +
xlab(NULL) + ylab('AUROC') +
scale_y_continuous(limits=c(ymin.auroc, 1), breaks=seq(0, 1, .2), expand=c(0.03, 0)) +
scale_colour_manual(values=pathophys.colors, name='Pathophysiology')
return(gg)
}
NAtoFALSE <- function(vec) {
vec[is.na(vec)] <- FALSE
return(vec)
}
AddPanelColumn <- function(auroc.df) {
auroc.df$panel <- 'Model'
auroc.df[NAtoFALSE(auroc.df$metric == 'Path Count'), 'panel'] <- 'Path Count'
auroc.df[NAtoFALSE(auroc.df$metric == 'DWPC (w=0.4)'), 'panel'] <- 'Degree-Weighted Path Count'
is.msig.auc <- NAtoFALSE(substr(auroc.df$name, 1, 1) == '{')
auroc.df[is.msig.auc, 'panel'] <- msig.strip.text
return(auroc.df)
}
MeanConfInt <- function(x) {t.test(x)$conf.int[1:2]}
PlotAUROCs <- function(auroc.df, perm.df) {
# Plots to graphics device
auroc.df <- AddPanelColumn(auroc.df)
ymin.auroc <- min(subset(auroc.df, auroc != 0)$auroc)
emdash.size <- 8
emdash.perm.size <- 6
jitter.width <- 0.3
point.size <- 1.75
if (! missing(perm.df)) {
perm.df <- AddPanelColumn(perm.df)
auroc.df$permuted <- FALSE
perm.df$permuted <- TRUE
auroc.df <- rbind(auroc.df, subset(perm.df, breadth == 'global'))
}
## MSigDB Plot
msig.df <- subset(auroc.df, panel == msig.strip.text | panel == 'Model')
msig.df$name <- gsub('[{}]', '', msig.df$name)
msig.disease.df <- subset(msig.df, breadth == 'disease_specific')
msig.global.df <- subset(msig.df, breadth == 'global')
# Order by global AUROC
if (! missing(perm.df)) {
msig.global.perm.df <- subset(msig.global.df, permuted)
msig.global.df <- subset(msig.global.df, ! permuted)
}
msig.levels <- as.character(msig.global.df$name[order(msig.global.df$auroc)])
msig.disease.df$name <- factor(msig.disease.df$name, levels=msig.levels)
msig.global.df$name <- factor(msig.global.df$name, levels=msig.levels)
if (! missing(perm.df)) {
msig.global.perm.df$name <- factor(msig.global.perm.df$name, levels=msig.levels)
}
set.seed(0); msig.plot <- ggplot(msig.disease.df, aes(name, auroc))
msig.plot <- PerfPlot(msig.plot, ymin.auroc) +
geom_point(data=msig.global.df, size=emdash.size, shape='\u2014', color=Solar('base02')) +
geom_point(aes(color=disease_pathophys), position=position_jitter(width=jitter.width),
alpha=0.7, size=point.size, show_guide=FALSE)
if (! missing(perm.df)) {
msig.plot <- msig.plot + geom_point(data=msig.global.perm.df,
size=emdash.perm.size, shape='\u2014', color=Solar('violet'))
}
## NonMSigDB Plot
nonmsig.df <- subset(auroc.df, panel != msig.strip.text)
nonmsig.df$panel <- factor(nonmsig.df$panel, levels=c('Degree-Weighted Path Count', 'Path Count', 'Model'))
nonmsig.disease.df <- subset(nonmsig.df, breadth == 'disease_specific' & panel != 'PCt')
nonmsig.global.df <- subset(nonmsig.df, breadth == 'global')
# Order by global AUROC
if (! missing(perm.df)) {
nonmsig.global.perm.df <- subset(nonmsig.global.df, permuted)
nonmsig.global.df <- subset(nonmsig.global.df, ! permuted)
}
nonmsig.levels <- unique(as.character(nonmsig.global.df$name[order(nonmsig.global.df$auroc)]))
nonmsig.disease.df$name <- factor(nonmsig.disease.df$name, levels=nonmsig.levels)
nonmsig.global.df$name <- factor(nonmsig.global.df$name, levels=nonmsig.levels)
if (! missing(perm.df)) {
nonmsig.global.perm.df$name <- factor(nonmsig.global.perm.df$name, levels=nonmsig.levels)
}
set.seed(0); nonmsig.plot <- ggplot(nonmsig.disease.df, aes(name, auroc))
nonmsig.plot <- PerfPlot(nonmsig.plot, ymin.auroc) +
geom_point(data=nonmsig.global.df, size=emdash.size, shape='\u2014', color=Solar('base02')) +
geom_point(aes(color=disease_pathophys), position=position_jitter(width=jitter.width),
alpha=0.7, size=point.size) +
theme(legend.key=element_rect(linetype='blank')) +
theme(legend.margin=grid::unit(1, 'points'))
if (! missing(perm.df)) {
nonmsig.plot <- nonmsig.plot + geom_point(data=nonmsig.global.perm.df,
size=emdash.perm.size, shape='\u2014', color=Solar('violet'))
}
gg = gridExtra::grid.arrange(msig.plot, nonmsig.plot, ncol=1, widths=c(1, 1))
return(gg)
}