-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathnostoc.R
283 lines (220 loc) · 12 KB
/
nostoc.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
## R script for the analysis of the transcriptomic data obtained using RNA-seq
## to characterized the role of a cryptic tRNA gene operon in the recovery from
## translational stress
## Authors: Ana B. Romero-Losada and Francisco J. Romero-Campero
## Contact: Francisco J. Romero-Campero - email: fran@us.es
## Load libraries
library(FactoMineR)
library(factoextra)
library(VennDiagram)
## Load gene expression data and convert it into a matrix
df.gene.expression <- read.table(file = "gene_expression_complete.tsv",sep="\t",as.is=T,header=T)
gene.expression <- as.matrix(df.gene.expression[,3:ncol(df.gene.expression)])
is.matrix(gene.expression)
head(gene.expression)
rownames(gene.expression) <- df.gene.expression$gene_name
## Compute mean gene expression
h0 <- (df.gene.expression$rep1_0 + df.gene.expression$rep2_0 + df.gene.expression$rep3_0)/3
h3 <- (df.gene.expression$rep1_3 + df.gene.expression$rep2_3 + df.gene.expression$rep3_3)/3
h9 <- (df.gene.expression$rep1_9 + df.gene.expression$rep2_9 + df.gene.expression$rep3_9)/3
h24 <- (df.gene.expression$rep1_24 + df.gene.expression$rep2_24 + df.gene.expression$rep3_24)/3
mean.gene.expression <- matrix(c(h0,h3,h9,h24),ncol=4)
colnames(mean.gene.expression) <- c("h0","h3","h9","h24")
rownames(mean.gene.expression) <- df.gene.expression$gene_name
## Perform Principal Component Analysis
pca.gene.expression <- data.frame(colnames(gene.expression),t(gene.expression))
colnames(pca.gene.expression)[1] <- "Sample"
res.pca <- PCA(pca.gene.expression, graph = FALSE,scale.unit = TRUE,quali.sup = 1 )
fviz_eig(res.pca, addlabels = TRUE, ylim = c(0, 70),main = "")
## Hierarchical clustering
res.hcpc <- HCPC(res.pca, graph=FALSE,nb.clust = 3)
## Generate representation of hierarchical tree
png(filename = "figures/hierarchical_tree.png",width = 800,height = 400)
fviz_dend(res.hcpc,k=4,
cex = 1, # Label size
palette = "jco", # Color palette see ?ggpubr::ggpar
rect = TRUE, rect_fill = TRUE, # Add rectangle around groups
rect_border = "jco", # Rectangle color
type="rectangle",
labels_track_height = 400 # Augment the room for labels
)
dev.off()
# Extract expressed genes
nrow(gene.expression)
gene.expression <- gene.expression[apply(X = gene.expression,MARGIN = 1,FUN = sum) != 0,]
nrow(gene.expression)
## Log2 transformation of gene expression data
log2.gene.expression <- log2(gene.expression + 1)
## Function for scatter plots comparing replicates
replicate.scatter.plot <- function(rep1,rep2,gexpress)
{
plot(x=gexpress[,rep1],y = gexpress[,rep2],pch=19,cex=0.8,col="darkgrey",xlab=paste(rep1,"h"),ylab=paste(rep2,"h"),cex.lab=1.5)
lines(x=c(-1,20),y=c(-1,20),lwd=2,col="red")
text(x = 4,y=14,labels = paste("cor =",paste(round(100*cor(gexpress[,rep1],gexpress[,rep2]),digits = 2),"%")),cex = 1.2)
}
replicate.scatter.plot(rep1 = "rep1_0", rep2 = "rep2_0", gexpress = log2.gene.expression)
replicate.scatter.plot(rep1 = "rep1_0", rep2 = "rep3_0", gexpress = log2.gene.expression)
replicate.scatter.plot(rep1 = "rep2_0", rep2 = "rep3_0", gexpress = log2.gene.expression)
replicate.scatter.plot(rep1 = "rep1_3", rep2 = "rep2_3", gexpress = log2.gene.expression)
replicate.scatter.plot(rep1 = "rep1_3", rep2 = "rep3_3", gexpress = log2.gene.expression)
replicate.scatter.plot(rep1 = "rep2_3", rep2 = "rep3_3", gexpress = log2.gene.expression)
replicate.scatter.plot(rep1 = "rep1_9", rep2 = "rep2_9", gexpress = log2.gene.expression)
replicate.scatter.plot(rep1 = "rep1_9", rep2 = "rep3_9", gexpress = log2.gene.expression)
replicate.scatter.plot(rep1 = "rep2_9", rep2 = "rep3_9", gexpress = log2.gene.expression)
replicate.scatter.plot(rep1 = "rep1_24", rep2 = "rep2_24", gexpress = log2.gene.expression)
replicate.scatter.plot(rep1 = "rep1_24", rep2 = "rep3_24", gexpress = log2.gene.expression)
replicate.scatter.plot(rep1 = "rep2_24", rep2 = "rep3_24", gexpress = log2.gene.expression)
## Differential expression analysis
library(limma)
## Specification of the experimental design
factor.experimental.design <- c(1,2,3,4,1,2,3,4,1,2,3,4)
limma.experimental.design <- model.matrix(~ -1+factor(factor.experimental.design))
colnames(limma.experimental.design) <- c("h0","h3","h9","h24")
## Linear model fit
linear.fit <- lmFit(log2.gene.expression, limma.experimental.design)
## Contrast specification and computation
contrast.matrix <- makeContrasts(h3-h0, h9-h0, h24-h0,
h9-h3, h24-h3,
h24-h9,
levels=c("h0","h3","h9","h24"))
contrast.linear.fit <- contrasts.fit(linear.fit, contrast.matrix)
contrast.results <- eBayes(contrast.linear.fit)
## Extract results
fc.threshold <- 2
q.val.threshold <- 0.05
final.degs <- c()
for(i in 1:6)
{
degs.results.i <- topTable(contrast.results, number=nrow(log2.gene.expression),coef=i,sort.by="logFC")
head(degs.results.i)
fold.change.i <- degs.results.i$logFC
p.values.i <- degs.results.i$P.Value
q.values.i <- degs.results.i$adj.P.Val
genes.ids.i <- rownames(degs.results.i)
activated.genes.i <- genes.ids.i[fold.change.i > log2(fc.threshold) & q.values.i < q.val.threshold]
repressed.genes.i <- genes.ids.i[fold.change.i < - log2(fc.threshold) & q.values.i < q.val.threshold]
final.degs <- c(final.degs,activated.genes.i,repressed.genes.i)
}
final.degs <- unique(final.degs)
length(final.degs)
degs.expression <- gene.expression[final.degs,]
## Co-expression Clustering analysis
## Load libraries and allow for multi-threading
library(cluster)
## Compute distance based on correlation for clustering analysis
gene.correlation <- cor(t(degs.expression))
distance.matrix <- 1 - gene.correlation
## Compute hierarchical clustering for different number of clusters
hierarchical.clustering <- hclust(as.dist(distance.matrix),method="average")
hclust.2 <- cutree(hierarchical.clustering,k=2)
hclust.3 <- cutree(hierarchical.clustering,k=3)
hclust.4 <- cutree(hierarchical.clustering,k=4)
hclust.5 <- cutree(hierarchical.clustering,k=5)
hclust.6 <- cutree(hierarchical.clustering,k=6)
hclust.7 <- cutree(hierarchical.clustering,k=7)
hclust.8 <- cutree(hierarchical.clustering,k=8)
hclust.9 <- cutree(hierarchical.clustering,k=9)
hclust.10 <- cutree(hierarchical.clustering,k=10)
## Compute PAM clustering for different number of clusters
pam.2 <- pam(as.dist(distance.matrix),k=2,diss=TRUE)
pam.3 <- pam(as.dist(distance.matrix),k=3,diss=TRUE)
pam.4 <- pam(as.dist(distance.matrix),k=4,diss=TRUE)
pam.5 <- pam(as.dist(distance.matrix),k=5,diss=TRUE)
pam.6 <- pam(as.dist(distance.matrix),k=6,diss=TRUE)
pam.7 <- pam(as.dist(distance.matrix),k=7,diss=TRUE)
pam.8 <- pam(as.dist(distance.matrix),k=8,diss=TRUE)
pam.9 <- pam(as.dist(distance.matrix),k=9,diss=TRUE)
pam.10 <- pam(as.dist(distance.matrix),k=10,diss=TRUE)
## Compute silhouette for hierarchical and PAM different clusterings
sil2 <- silhouette(hclust.2,dist=distance.matrix)
sil3 <- silhouette(hclust.3,dist=distance.matrix)
sil4 <- silhouette(hclust.4,dist=distance.matrix)
sil5 <- silhouette(hclust.5,dist=distance.matrix)
sil6 <- silhouette(hclust.6,dist=distance.matrix)
sil7 <- silhouette(hclust.7,dist=distance.matrix)
sil8 <- silhouette(hclust.8,dist=distance.matrix)
sil9 <- silhouette(hclust.9,dist=distance.matrix)
sil10 <- silhouette(hclust.10,dist=distance.matrix)
hclust.sil.values <- c(summary(sil2)[["avg.width"]],summary(sil3)[["avg.width"]],summary(sil4)[["avg.width"]],summary(sil5)[["avg.width"]],summary(sil6)[["avg.width"]],summary(sil7)[["avg.width"]],summary(sil8)[["avg.width"]],summary(sil9)[["avg.width"]],summary(sil10)[["avg.width"]])
sil2 <- silhouette(pam.2)
sil3 <- silhouette(pam.3)
sil4 <- silhouette(pam.4)
sil5 <- silhouette(pam.5)
sil6 <- silhouette(pam.6)
sil7 <- silhouette(pam.7)
sil8 <- silhouette(pam.8)
sil9 <- silhouette(pam.9)
sil10 <- silhouette(pam.10)
pam.sil.values <- c(summary(sil2)[["avg.width"]],summary(sil3)[["avg.width"]],summary(sil4)[["avg.width"]],summary(sil5)[["avg.width"]],summary(sil6)[["avg.width"]],summary(sil7)[["avg.width"]],summary(sil8)[["avg.width"]],summary(sil9)[["avg.width"]],summary(sil10)[["avg.width"]])
## Plot silhouettes
plot(2:10,pam.sil.values,type="o",col="blue",pch=0,ylim=c(0.3,0.8),xlab="Number of clusters",ylab="Silhouette",lwd=3)
lines(2:10,hclust.sil.values,type="o",col="red",pch=1,xlab="",ylab="",lwd=3)
legend("topright",legend=c("PAM","HCLUST"),col=c("blue","red"),pch=c(0,1),lwd=3)
plot(2:6,pam.sil.values[1:5],type="o",col="blue",pch=0,ylim=c(0.3,0.8),xlab="Number of clusters",ylab="Silhouette",lwd=3)
lines(2:6,hclust.sil.values[1:5],type="o",col="red",pch=1,xlab="",ylab="",lwd=3)
legend("topright",legend=c("PAM","HCLUST"),col=c("blue","red"),pch=c(0,1),lwd=3)
## Extract best clutering PAM with Three clustres
cluster1.pam3 <- names(which(pam.3[["clustering"]] == 1))
length(cluster1.pam3)
cluster2.pam3 <- names(which(pam.3[["clustering"]] == 2))
length(cluster2.pam3)
cluster3.pam3 <- names(which(pam.3[["clustering"]] == 3))
length(cluster3.pam3)
## Extract gene expression for each cluster
expr.cluster1.pam3 <- mean.gene.expression[cluster1.pam3,]
expr.cluster2.pam3 <- mean.gene.expression[cluster2.pam3,]
expr.cluster3.pam3 <- mean.gene.expression[cluster3.pam3,]
## compute average cluster gene expression
mean.profile.cluster1.pam3 <- colMeans(expr.cluster1.pam3)
mean.profile.cluster2.pam3 <- colMeans(expr.cluster2.pam3)
mean.profile.cluster3.pam3 <- colMeans(expr.cluster3.pam3)
## Plot average cluster gene expression
plot(x = c(0,3,9,24),mean.profile.cluster1.pam3,type="o",col="blue",xlab="Time (h)",ylab="Expression Level (FPKM)",lwd=3,pch=0,main="Cluster Profiles",ylim=c(0,800))
lines(x = c(0,3,9,24),mean.profile.cluster2.pam3,type="o",col="red",lwd=3,pch=1)
lines(x = c(0,3,9,24),mean.profile.cluster3.pam3,type="o",col="darkgreen",lwd=3,pch=1)
## Write genes for each cluster
write.table(x = cluster1.pam3, file = "cluster_blue_fast_activation.txt",quote = F,row.names = F,col.names = F)
write.table(x = cluster2.pam3, file = "cluster_red_steady_activation.txt",quote = F,row.names = F,col.names = F)
write.table(x = cluster3.pam3, file = "cluster_green_repression.txt",quote = F,row.names = F,col.names = F)
## kEGG pathway enrichment analysis
library(clusterProfiler)
kk.1 <- enrichKEGG(gene = cluster1.pam3,
organism = 'ana',
pvalueCutoff = 0.05)
as.data.frame(kk.1)
kk.2 <- enrichKEGG(gene = cluster2.pam3,
organism = 'ana',
pvalueCutoff = 0.05)
as.data.frame(kk.2)
kk.3 <- enrichKEGG(gene = cluster3.pam3,
organism = 'ana',
pvalueCutoff = 0.05)
as.data.frame(kk.3)
## Heatmaps
head(mean.gene.expression)
dim(mean.gene.expression)
clusters.mean.expression <- mean.gene.expression[c(cluster1.pam3,cluster2.pam3,cluster3.pam3),]
genes <- rownames(clusters.mean.expression)
scaled.gene.expression <- matrix(nrow=length(genes),ncol=97)
for(i in 1:length(genes))
{
current.gene <- genes[i]
expression.current.gene.i <- clusters.mean.expression[current.gene,]/max(clusters.mean.expression[current.gene,])
scaled.gene.expression[i,] <- approx(x=c(0,3,9,24),expression.current.gene.i,xout=seq(from=0,to=24,by=0.25))[[2]]
}
library(gplots)
colfunc <- colorRampPalette(c("black","blue","yellow"))
png(filename = "figures/heatmap.png",width = 600)
heatmap.2(scaled.gene.expression,Colv =FALSE,trace=c("none"),dendrogram = "none",col=colfunc(100),density.info=c("none"),key=FALSE,margins = c(2,8),cexRow = 1.2,Rowv=FALSE,labCol = c(""),labRow = "")
dev.off()
scaled.gene.expression <- matrix(nrow=length(genes),ncol=4)
for(i in 1:length(genes))
{
current.gene <- genes[i]
scaled.gene.expression[i,] <- clusters.mean.expression[current.gene,]/max(clusters.mean.expression[current.gene,])
}
library(gplots)
colfunc <- colorRampPalette(c("black","blue","yellow"))
png(filename = "figures/heatmap_no_interpolation.png",width = 600)
heatmap.2(scaled.gene.expression,Colv =FALSE,trace=c("none"),dendrogram = "none",col=colfunc(100),density.info=c("none"),key=FALSE,margins = c(2,8),cexRow = 1.2,Rowv=FALSE,labCol = c(""),labRow = "")
dev.off()