-
Notifications
You must be signed in to change notification settings - Fork 0
/
Kmeans_pam.R
133 lines (108 loc) · 4.54 KB
/
Kmeans_pam.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
rm(list=ls())
library(RColorBrewer)
library(cluster)
library(reshape2)
library(dplyr)
library(ggplot2)
setwd('C:\\Users\\R')
### read expression data
exprSet<-read.delim('All_gene_fpkm.list',header = T,
#row.names = 1,
sep='\t',comment.char = '',
quote = '',check.names = F)[1:2000,]
### read duplicated information
group_frame<-read.delim('group_list.txt',sep='\t',header = F,stringsAsFactors = F)
'''
V1 V2
ApEpC A
ApHeC A
ApEpY B
ApMgY B
ApHeY C
ApMgC C
'''
sample_names<-colnames(exprSet)
newdataframe<-exprSet[match(colnames(exprSet),as.character(group_frame[,1])),]
match_samples<-colnames(exprSet)[2:length(colnames(exprSet))][match(group_frame[,1],sample_names[2:length(sample_names)])] #return index, last subjected to first arrange
### sort column with previously vector match_samples
new_exprSet<-exprSet[
,with(exprSet, match_samples)
]
library(tibble) # including has_rownames function
has_rownames(new_exprSet)
new_exprSet<-remove_rownames(new_exprSet)
new_exprSet$ID<-exprSet[,1]
temp1 <- new_exprSet[, c(length(new_exprSet), 1:length(new_exprSet)-1)]
temp_group<-table(group_frame$V2)
getmeancol<-function(data_frame,ingroup,tab_group){
j<-0
new_data_frame<-data.frame(ID=data_frame[,1])
for(i in unique(ingroup$V2)){
j =j+as.numeric(tab_group[[i]])
new_data_frame[,i]<-rowMeans(select(data_frame,j:j+1))
}
return(new_data_frame)
}
new_means_datafram<-getmeancol(temp1,group_frame,temp_group)
new_means_datafram<-exprSet
#colnames(exprSet)<-sub('(#ID|ID)','Gene',names(exprSet))
#exprSet_tibble<-as.tibble(exprSet) #转换为dplyr的tibble数据框
names(exprSet)[1]<-'Gene'
samplename<-1:(length(colnames(exprSet))-1)
names(samplename)<-colnames(exprSet)[2:length(colnames(exprSet))]
str(samplename)
#names(samplename)<-colnames(exprSet)[which(colnames(exprSet)!='ID')]
for(i in samplename){
colnames(exprSet)[i+1]=i
}
rownames(exprSet)<-exprSet[,1]
exprSet1<-exprSet %>%select(2:7) #select 2-7heatmap.2(transposed_alpha_mtx,
exprSet_t<-as.data.frame(t(exprSet1))
exprSet_cor<-exprSet_t %>%
cor(use='pairwise.complete.obs',method='pearson') #Obtain cor list，pairwise.complete.obs was parameters needed Miss values(NA)
### draw cor heatmap throught heatmap.2 function within gplots
if(!requireNamespace('gplots')) install.packages('gplots')
library(gplots) #including heatmap.2 function
color_scheme <- rev(brewer.pal(8,"RdBu"))
pdf(file='heatmap.2.pdf',height = 10,width=8)
heatmap.2(exprSet_cor[1:500,1:500],na.rm=TRUE,
cexRow=0.5, cexCol=0.5,
Rowv = NULL, # use the dendrogram previously calculated
Colv = NULL, # don't mess with my columns! (keep current ordering )
dendrogram = NULL, # only draw row dendrograms
breaks = seq(-3, 3, length.out = 9), # OPTIONAL: set break points for colors
col = color_scheme, # use previously defined colors
trace = "none", density.info = "none" # remove distracting elements of plot
)
dev.off()
exprSet_dist <- as.dist(1 - exprSet_cor) #deleting positive correlation from cor matrix by as.dist
fpkm_kmedoids <- pam(exprSet_dist, 8)
k_pam_clusters <- fpkm_kmedoids$cluster
sum(table(k_pam_clusters))
clusters_df <- data.frame(Gene = names(k_pam_clusters),
cluster = as.factor(k_pam_clusters))
exprSet_t$ID<-rownames(exprSet_t)
exprSet_m<-melt(exprSet_t,id.vars="ID",variable.name="Gene",value.name="expression")
exprSet_m$ID<-as.numeric(exprSet_m$ID)
exprSet_m<-exprSet_m %>%
left_join(clusters_df,by=c('Gene')) #left join the cluster_df into exprSet_m
cluster_means<-exprSet_m %>% group_by(cluster,ID) %>%
summarize(mean.exp=mean(expression,na.rm=TRUE))
options(warn=-1) #suppressing warning
p<-exprSet_m %>%
ggplot(aes(ID, expression, group=Gene)) +
geom_line(alpha=0.25) +
geom_line(aes(ID, mean.exp, group=NULL,color=cluster),
data = cluster_means,
size=1.1) +
ylim(0, 2) + #y tick arrange
facet_wrap(~cluster, ncol=4)+
theme_bw()+
scale_x_discrete(limits=c(names(samplename)))+
xlab('SampleName')+
theme(axis.text.x = element_text(face="bold", color="black",
size=8, angle=45,margin=margin(8,0,0,0)), #margin funs adjustting space between x axis and x text
axis.text.y=element_text(face="bold", color="black",size=8))
ggsave(p,filename = 'K-medoids_Cluster.pdf',path='./',width=22,height=15,units='cm',colormodel='srgb')
?ggsave
### Done