-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathPCA_PCoA_PLSDA_UPGMA.R
269 lines (231 loc) · 12 KB
/
PCA_PCoA_PLSDA_UPGMA.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
#! /urs/bin/env Rscript
## ---------------------------
##
## Script name: PCA_PCoA_PLSDA_UPGMA.R
##
## Purpose of script:计算基于otu表的PCA\PCoA\PLSDA\CA\UPGMA
##
## Author: liuc
##
## Date Created: 2020-02-21
##
## Copyright (c)
## Email:
##
## ---------------------------
##
## Notes: 每次只传入一组数据，所以要对map表进行筛选
## Usage: Rscript
##
## ---------------------------
options(BioC_mirror='http://mirrors.ustc.edu.cn/bioc/')
#options(BioC_mirror="https://mirrors.tuna.tsinghua.edu.cn/bioconductor")
args <- commandArgs(trailingOnly = TRUE)
stopifnot(length(args) >= 3)
infile <- args[1]
map_file <- args[2]
outpath <- args[3]
group_str <- args[4]
one_group <- eval(parse(text = (group_str)))
print(one_group)
#library(plyr)
library(tidyverse)
library(vegan)
library(ropls)
library(mixOmics)
library(dendextend)
library(RColorBrewer)
library(ggrepel)
tax_color <- c('#8DD3C7', '#FFFFB3', '#BEBADA', '#FB8072', '#80B1D3',
'#FDB462', '#B3DE69', '#FCCDE5', '#BC80BD', '#CCEBC5', 'gray')
# infile <- '/Users/congliu/prs/ZJDXGW/combined_otu_table_m2_std.txt'
# map_file <- '/Users/congliu/prs/ZJDXGW/map.txt'
# outpath <- '/Users/congliu/prs/ZJDXGW/'
# one_group <- c('A', 'B', 'C', 'D','E')
group1 <- one_group[1]
group2 <- one_group[2]
# group1 <- 'all'
# group2 <- 'group'
otu <- read.delim(infile, row.names = 1, sep = '\t',
stringsAsFactors = FALSE, check.names = FALSE)
mapfile <- read.delim(map_file, sep = '\t') %>% dplyr::select(Type, Description)
otu <- subset(otu, taxonomy != 'Unassigned')
otu <- subset(otu, select = -taxonomy)
otu <- data.frame(t(otu))
otu2 <- log(otu + 1)
#otu2 <- otu
otu2$Description <- row.names(otu2)
otu3 <- merge(otu2, mapfile, by.x = 'Description', by.y = 'Description')
row.names(otu3) <- otu3$Description
otu3 <- subset(otu3, Type %in% one_group) ##挑选出一组样本，提供分组信息
otu4 <- subset(otu3, select = -c(Description, Type)) #为所有计算所用的otu数值
#############PCoA#################
# I have used the weighted UniFrac metric to determine the distance between samples and PCoA to visualise the data
# 但是下文不是
distance <- vegan::vegdist(otu4, method = 'bray')
pcoa <- cmdscale(distance, k = (nrow(otu4) - 1), eig = TRUE)
#write.csv(as.matrix(distance), 'distance.csv', quote = F)
#ordiplot(scores(pcoa)[ ,c(1, 2)], type = 't')
species <- wascores(pcoa$points[,1:2], otu4)
pcoa_eig <- (pcoa$eig)[1:2] / sum(pcoa$eig)
sample_site <- data.frame({pcoa$point})[1:2]
sample_site$names <- rownames(sample_site)
names(sample_site)[1:2] <- c('PCoA1', 'PCoA2')
sample_site <- merge(sample_site, mapfile, by.x = 'names', by.y = 'Description', all.x = TRUE)
group_border <- plyr::ddply(sample_site, 'Type', function(df) df[chull(df[[2]], df[[3]]), ])
p_pcoa <- ggplot(sample_site, aes(PCoA1, PCoA2, group = Type)) +
theme(panel.grid = element_line(color = 'gray', linetype = 2, size = 0.1),
panel.background = element_rect(color = 'black', fill = 'transparent'),
legend.key = element_rect(fill = 'transparent')) +
geom_vline(xintercept = 0, color = 'gray', size = 0.4) +
geom_hline(yintercept = 0, color = 'gray', size = 0.4) +
geom_polygon(data = group_border, aes(fill = Type)) +
geom_point(aes(color = Type), size = 1.5, alpha = 0.8) +
scale_shape_manual(values = c(17, 16)) +
scale_color_manual(values = c('orange', 'blue2', 'red4', '#C673FF2E', 'green2')) +
scale_fill_manual(values = c('#C673FF2E', '#73D5FF2E', '#49C35A2E', '#FF985C2E', '#C673FF2E')) + #可在这里修改区块的颜色
guides(fill = guide_legend(order = 1), shape = guide_legend(order = 2), color = guide_legend(order = 3)) +
labs(x = paste('PCoA axis1: ', round(100 * pcoa_eig[1], 2), '%'),
y = paste('PCoA axis2: ', round(100 * pcoa_eig[2], 2), '%'))
#可通过修改下面四句中的点坐标、大小、颜色等，修改“A、B、C、D”标签
# annotate('text', label = 'A', x = -0.31, y = -0.15, size = 5, colour = '#C673FF') +
# annotate('text', label = 'B', x = -0.1, y = 0.3, size = 5, colour = '#73D5FF') +
# annotate('text', label = 'C', x = 0.1, y = 0.15, size = 5, colour = '#49C35A') +
# annotate('text', label = 'D', x = 0.35, y = 0, size = 5, colour = '#FF985C')
ggsave(p_pcoa, filename = file.path(outpath, paste('pcoa_', group1, '_', group2, '.pdf', sep = '')), height = 8, width = 8)
#####################PCA####################
#Transformation-based Principal Component Analysis，tb-PCA
#因为PCA计算的欧几里得距离的原因, 而物种数据0较多，可选择进行hellinger，hellinger转化本身会降低高丰度物种的权重
otu5 <- otu
otu5$Description <- row.names(otu5)
otu6 <- merge(otu5, mapfile, by.x = 'Description', by.y = 'Description')
row.names(otu6) <- otu6$Description
otu6 <- subset(otu6, Type %in% one_group)
otu7 <- subset(otu6, select = -c(Description, Type))
otu_hel <- decostand(otu7, method = 'hellinger')
# rda scale为TRUE时排序对象是相关矩阵、scale为FASLE时排序对象是离差矩阵
pca_sp <- rda(otu_hel, scale = FALSE)
#ordiplot(pca_sp, scaling = 1, display = 'site', type = 'points')
pca_eig <- pca_sp$CA$eig
pca_exp <- pca_sp$CA$eig / sum(pca_sp$CA$eig)
#site.scaling1 <- scores(pca_exp, choices = 1:2, scaling = 1, display = 'site')
#env.scaling2 <- scores(pca_exp, choices = 1:2, scaling = 2, display = 'sp')
site.scaling1 <- data.frame(summary(pca_sp, scaling = 1)$sites[ ,1:2])
otu.scaling1 <- data.frame(summary(pca_sp, scaling = 1)$species[ ,1:2])
site.scaling1$sample <- rownames(site.scaling1)
site.scaling1 <- merge(site.scaling1, mapfile, by.x = 'sample', by.y = 'Description')
otu.scaling1$group <- rownames(otu.scaling1)
otu.scaling1 <- otu.scaling1[1:5,]
pc1 <- paste('PC1:', round(pca_exp[1]*100, 2), '%')
pc2 <- paste('PC2:',round(pca_exp[2]*100, 2), '%')
p_pca <- ggplot(site.scaling1, aes(PC1, PC2)) +
geom_point(aes(color = Type)) +
stat_ellipse(aes(PC1, PC2, color = Type), show.legend = FALSE, linetype = 2) +
scale_color_manual(values = c('red4', 'green3', '#C673FF2E', 'orange', 'blue2')) +
scale_shape_manual(values = c(16, 17)) +
theme(panel.grid = element_blank(), panel.background = element_rect(color = 'black', fill = 'transparent'),
legend.title = element_blank(), legend.key = element_rect(fill = 'transparent')) +
labs(x = pc1, y = pc2) +
geom_vline(xintercept = 0, color = 'gray', size = 0.5) +
geom_hline(yintercept = 0, color = 'gray', size = 0.5) +
geom_segment(data = otu.scaling1, aes(x = 0, y = 0, xend = PC1, yend = PC2),
arrow = arrow(length = unit(0.1, 'cm')), size = 0.3, color = 'blue') +
geom_text(data = otu.scaling1, aes(PC1 * 1.1, PC2 * 1.1, label = group), color = 'blue', size = 3)
ggsave(p_pca, filename = file.path(outpath, paste('pca_', group1, '_', group2, '.pdf', sep = '')),
height = 8, width = 8)
##################PLSDA######################
# otu.plsda <- opls(x = otu4, y = factor(otu3$Type), orthoI = 0)
# head(otu.plsda@scoreMN)
# plot(otu.plsda, typeVc = 'overview', parAsColFcVn = factor(otu3$Type))
plsda_result <-plsda(otu4, otu3$Type, ncomp = 2)
#plotIndiv(plsda_result, ind.names = TRUE, style = 'ggplot2')
sample_site <- data.frame(plsda_result$variates)[1:2]
sample_site$names <- rownames(sample_site)
names(sample_site)[1:2] <- c('plsda1', 'plsda2')
sample_site <- merge(sample_site, mapfile, by.x = 'names', by.y = 'Description', all.x = TRUE)
plsda_result_eig <- {plsda_result$explained_variance$X}[1:2]
p_plsda <- ggplot(sample_site, aes(plsda1, plsda2, color = Type, label = names)) +
geom_point(size = 1.5, alpha = 0.6) +
stat_ellipse(show.legend = FALSE, linetype = 2) +
scale_color_manual(values = c('#1D7ACC', '#F67433', '#00815F','#C673FF2E', 'blue2')) +
scale_shape_manual(values = c(16, 17)) +
theme(panel.grid = element_line(color = 'grey50'),
panel.background = element_rect(color = 'black', fill = 'transparent')) +
theme(legend.title = element_blank(), legend.key = element_rect(fill = 'transparent')) +
labs(x = paste('PLS-DA axis1 ( explained variance ', round(100 * plsda_result_eig[1], 2), '% )', sep = ''),
y = paste('PLS-DA axis2 ( explained variance ', round(100 * plsda_result_eig[2], 2), '% )', sep = ''))
ggsave(p_plsda, filename = file.path(outpath, paste('plsda_', group1, '_', group2, '.pdf', sep = '')), height = 8, width = 8)
###################CA/DCA/CCA/RDA####################
otu5 <- otu
otu5$Description <- row.names(otu5)
otu6 <- merge(otu5, mapfile, by.x = 'Description', by.y = 'Description')
row.names(otu6) <- otu6$Description
otu6 <- subset(otu6, Type %in% one_group) ##挑选出一组样本，提供分组信息
otu7 <- subset(otu6, select = -c(Description, Type)) #为所有计算所用的otu数值
ca <- vegan::cca(otu7)
ca_eig <- ca$CA$eig
ca_exp <- ca$CA$eig / sum(ca$CA$eig)
ca.scaling1 <- data.frame(summary(ca, scaling = 1)$sites[ ,1:2])
otu.scaling1 <- data.frame(summary(ca, scaling = 1)$species[ ,1:2])
ca.scaling1$sample <- rownames(ca.scaling1)
ca.scaling1 <- merge(ca.scaling1, mapfile, by.x = 'sample', by.y = 'Description', all.x = TRUE)
otu.scaling1$group <- rownames(otu.scaling1)
#只保留 top10 丰度的细菌门
top10_otu <- names(colSums(otu7)[order(colSums(otu7), decreasing = TRUE)][1:10])
top10_otu.scaling1 <- otu.scaling1[top10_otu, ]
p_ca <- ggplot(ca.scaling1, aes(CA1, CA2)) +
geom_point(aes(color = Type)) +
scale_color_manual(values = c('red', 'blue2')) +
scale_shape_manual(values = c(16, 17)) +
theme(panel.grid = element_blank(), panel.background = element_rect(color = 'black', fill = 'transparent'),
legend.key = element_rect(fill = 'transparent'), plot.title = element_text(hjust = 0.5)) +
geom_vline(xintercept = 0, color = 'gray', size = 0.5) +
geom_hline(yintercept = 0, color = 'gray', size = 0.5) +
geom_point(data = top10_otu.scaling1, color = 'green') +
geom_text_repel(data = top10_otu.scaling1, aes(label = group),
size = 3, box.padding = unit(0.5, 'lines'), color = 'blue')
#labs(x = ca1, y = ca2, title = 'I 型标尺，仅展示 top10 丰度种群', color = NULL)
ggsave(p_ca, filename = file.path(outpath, paste('ca_', group1, '_', group2, '.pdf', sep = '')), height = 8, width = 8)
##################UPGMA##################
distance <- vegan::vegdist(otu4, method = 'bray')
dis_ch <- vegdist(decostand(otu4, 'normalize'), method = 'euc') #弦距离，权重的？
#使用中值的非权重成对组法，UPGMA
clust_average <- hclust(distance, method = 'average')
#使用质心的非权重成对组法，UPGMC
clust_centroid <- hclust(distance, method = 'centroid')
#使用中值的权重成对组法，WPGMA
clust_mcquitty <- hclust(distance, method = 'mcquitty')
#使用质心的权重成对组法，WPGMC
clust_median <- hclust(distance, method = 'median')
##Ward 最小方差聚类（包含两种不同的 Ward 聚类算法，详情 ?hclust）
#该方法中只能使用欧式几何属性的距离测度，因此更换为平方根转化后的 Bray-curtis 距离测度
dis_bray_euc <- distance^0.5
#ward.D
clust_ward <- hclust(dis_bray_euc, method = 'ward.D')
#ward.D2
clust_ward2 <- hclust(dis_bray_euc, method = 'ward.D2')
clust_average <- hclust(distance, method = 'average')
#plot(clust_average, main = 'UPGMA\nBray-curtis', sub = '', xlab = 'Sample', ylab = 'Height')
#rect.hclust(clust_average, k = 2, border = c('red', 'blue'))
hcd=as.dendrogram(clust_average)
labelColors=brewer.pal(n=4, name="Set1")
clusMember <- otu3$Type
names(clusMember) <- otu3$Description
#clusMember=cutree(hcd, 4)
hcd=hcd %>% set("labels_cex", 1.5) %>% set("branches_lwd", 2) %>%
set("branches_k_color", k=4) %>% set("branches_k_lty", k=4)
colLab <- function(n) {
if (is.leaf(n)) {
a <- attributes(n)
labCol <- labelColors[clusMember[which(names(clusMember) == a$label)]]
attr(n, 'nodePar') <- c(a$nodePar, lab.col = labCol)
}
n
}
clusDendro=dendrapply(hcd, colLab)
# pdf(file.path(outpath, paste('upgma_', group1, '_', group2, '.pdf', sep = '')),
# width = 800, height = 800)
#plot(clusDendro, main="UPGMA Tree", type="rectangle", horiz=FALSE, ylab = 'Height', cex=0.1)
plot(clusDendro, main = 'UPGMA\n(Bray-curtis distance)', sub = '', ylab = 'Height')
#axis(side=1, cex.lab=0.1)
dev.off()