-
Notifications
You must be signed in to change notification settings - Fork 0
/
UMAP-CONTROL--TimePoint_1 + TimePoint_2
321 lines (280 loc) · 17.3 KB
/
UMAP-CONTROL--TimePoint_1 + TimePoint_2
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
# R version 4.1.3(x64 bit) and RStudio 2022.07.1+554 "Spotted Wakerobin" Release (7872775ebddc40635780ca1ed238934c3345c5de, 2022-07-22) for Windows
Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) QtWebEngine/5.12.8 Chrome/69.0.3497.128 Safari/537.36 were used for running this code :)
# Seurat is a multimodal single Cell RNA seq analysis algorithm created by
# The Satija Lab. For more information please see: https://satijalab.org/seurat/
# LOAD LIBRARIES ####
# Restart Rstudio or R
install.packages("devtools")
install.packages("usethis")
library(usethis)
library(devtools)
install.packages("pkgbuild")
library(pkgbuild)
install.packages("Matrix")
install.packages("ggridges")
install.packages("cowplot")
install.packages('ggrepel')
install.packages("R.utils")
install.packages("gridExtra")
install.packages("Seurat")
install.packages("plotly")
install.packages("clustree")
install.packages('multtest')
install.packages("ggplot2")
install.packages("ggraph")
install.packages('Seurat')
if (!requireNamespace("BiocManager", quietly = TRUE))
install.packages("BiocManager")
BiocManager::install("monocle")
# Install multtest for Seurat
BiocManager::install("multtest")
library(ggplot2)
library(cowplot)
library(Matrix)
library(ggridges)
library(ggrepel)
library(dplyr)
library(Seurat)
library(plotly)
library(clustree)
library(Seurat)
library(ggraph)
library(Matrix)
library(VGAM)
library(stats4)
library(splines)
library(DDRTree)
library(irlba)
library(BiocGenerics)
library(Biobase)
# Run the following code once you have Seurat installed
# Set global environment parameter
options(future.globals.maxSize= 89128960)
memory.limit()
memory.limit(size=21474836480)
# Setup CONTROL-ST2+ST3(Timepoint 1 and TimePoint 2) Seuratobject
# OBJECT SETUP AND NORMALIZATION ####
# Load 10X data ####
# Create Seurat objects ####
ctrl1st2.data <-Read10X(data.dir = "C:/Users/mad1188/Box/Sequencing Data/cellranger_count/ST2-CTRL/filtered_feature_bc_matrix")
ctrl1st2 <- CreateSeuratObject(counts = ctrl1st2.data,
project = "ctrl1st2"
)
ctrl2st2.data <-Read10X(data.dir = "C:/Users/mad1188/Box/Sequencing Data/cellranger_count/ST-2-CTRL-1/filtered_feature_bc_matrix")
ctrl2st2 <- CreateSeuratObject(counts = ctrl2st2.data,
project = "ctrl2st2")
ctrl3st2.data <-Read10X(data.dir = "C:/Users/mad1188/Box/Sequencing Data/cellranger_count/ST-2-CTRL-2/filtered_feature_bc_matrix")
ctrl3st2 <- CreateSeuratObject(counts = ctrl3st2.data,
project = "ctrl3st2"
)
head(x = colnames(x = ctrl1st2))
ctrl1st2 <- RenameCells(object = ctrl1st2, add.cell.id = "ctrl1st2")
head(x = colnames(x = ctrl1st2))
head(x = colnames(x = ctrl2st2))
ctrl2st2 <- RenameCells(object = ctrl2st2, add.cell.id = "ctrl2st2")
head(x = colnames(x = ctrl2st2))
head(x = colnames(x = ctrl3st2))
ctrl3st2 <- RenameCells(object = ctrl3st2, add.cell.id = "ctrl3st2")
head(x = colnames(x = ctrl3st2))
ctrl1st2[["percent.mt"]] <- PercentageFeatureSet(object = ctrl1st2, pattern = "^MT-")
ctrl2st2[["percent.mt"]] <- PercentageFeatureSet(object = ctrl2st2, pattern = "^MT-")
ctrl3st2[["percent.mt"]] <- PercentageFeatureSet(object = ctrl3st2, pattern = "^MT-")
p101 <- VlnPlot(object = ctrl1st2, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3)
p201 <- VlnPlot(object = ctrl2st2, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3)
p301 <- VlnPlot(object = ctrl3st2, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3)
CombinePlots(plots = list(p101, p201, p301), ncol =1)
p401 <- FeatureScatter(object = ctrl1st2, feature1 = "nCount_RNA", feature2 = "percent.mt")
p501 <- FeatureScatter(object = ctrl1st2, feature1 = "nCount_RNA", feature2 = "nFeature_RNA")
p601 <- FeatureScatter(object = ctrl2st2, feature1 = "nCount_RNA", feature2 = "percent.mt")
p701 <- FeatureScatter(object = ctrl2st2, feature1 = "nCount_RNA", feature2 = "nFeature_RNA")
p801 <- FeatureScatter(object = ctrl3st2, feature1 = "nCount_RNA", feature2 = "percent.mt")
p901 <- FeatureScatter(object = ctrl3st2, feature1 = "nCount_RNA", feature2 = "nFeature_RNA")
CombinePlots(plots = list(p401, p501, p601, p701, p801, p901), ncol =2)
# Thresholding ####
# RNA based cell thresholding
# The operator can add columns to object metadata. This is a great place to stash QC stats
ctrl1st2 <- subset(x = ctrl1st2, subset = nFeature_RNA > 200 & nFeature_RNA < 8000 & percent.mt < 20)
ctrl2st2 <- subset(x = ctrl2st2, subset = nFeature_RNA > 200 & nFeature_RNA < 8000 & percent.mt < 20)
ctrl3st2 <- subset(x = ctrl3st2, subset = nFeature_RNA > 200 & nFeature_RNA < 8000 & percent.mt < 20)
p1001 <- VlnPlot(object = ctrl1st2, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3)
p1101 <- VlnPlot(object = ctrl2st2, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3)
p1201 <- VlnPlot(object = ctrl3st2, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3)
CombinePlots(plots = list(p1001, p1101, p1201), ncol =1)
p1301 <- FeatureScatter(object = ctrl1st2, feature1 = "nCount_RNA", feature2 = "percent.mt")
p1401 <- FeatureScatter(object = ctrl1st2, feature1 = "nCount_RNA", feature2 = "nFeature_RNA")
p1501 <- FeatureScatter(object = ctrl2st2, feature1 = "nCount_RNA", feature2 = "percent.mt")
p1601 <- FeatureScatter(object = ctrl2st2, feature1 = "nCount_RNA", feature2 = "nFeature_RNA")
p1701 <- FeatureScatter(object = ctrl3st2, feature1 = "nCount_RNA", feature2 = "percent.mt")
p1801 <- FeatureScatter(object = ctrl3st2, feature1 = "nCount_RNA", feature2 = "nFeature_RNA")
CombinePlots(plots = list(p1301, p1401, p1501, p1601, p1701, p1801), ncol =2)
ctrlST2 <- sample(x = c("ctrlST2"), size = ncol(x = ctrl1st2), replace = TRUE)
ctrl1st2$groups <- ctrlST2
head(x = ctrl1st2[[]])
ctrlST2 <- sample(x = c("ctrlST2"), size = ncol(x = ctrl2st2), replace = TRUE)
ctrl2st2$groups <- ctrlST2
head(x = ctrl2st2[[]])
ctrlST2 <- sample(x = c("ctrlST2"), size = ncol(x = ctrl3st2), replace = TRUE)
ctrl3st2$groups <- ctrlST2
head(x = ctrl3st2[[]])
# OBJECT SETUP AND NORMALIZATION ####
# Load 10X data ####
# Create Seurat objects ####
control1st3.data <-Read10X(data.dir = "C:/Users/mad1188/Box/Sequencing Data/cellranger_count/ST-3-CTRL-1/filtered_feature_bc_matrix")
control2st3.data <-Read10X(data.dir = "C:/Users/mad1188/Box/Sequencing Data/cellranger_count/ST-3-CTRL-2/filtered_feature_bc_matrix")
control3st3.data <-Read10X(data.dir = "C:/Users/mad1188/Box/Sequencing Data/cellranger_count/ST-3-CTRL-3/filtered_feature_bc_matrix")
control1st3 <- CreateSeuratObject(counts = control1st3.data,
project = "control1st3"
)
control2st3 <- CreateSeuratObject(counts = control2st3.data,
project = "control2st3"
)
control3st3 <- CreateSeuratObject(counts = control3st3.data,
project = "control3st3"
)
head(x = colnames(x = control1st3))
control1st3 <- RenameCells(object = control1st3, add.cell.id = "control1st3")
head(x = colnames(x = control1st3))
head(x = colnames(x = control2st3))
control2st3 <- RenameCells(object = control2st3, add.cell.id = "control2st3")
head(x = colnames(x = control2st3))
head(x = colnames(x = control3st3))
control3st3 <- RenameCells(object = control3st3, add.cell.id = "control3st3")
head(x = colnames(x = control3st3))
control1st3[["percent.mt"]] <- PercentageFeatureSet(object = control1st3, pattern = "^MT-")
control2st3[["percent.mt"]] <- PercentageFeatureSet(object = control2st3, pattern = "^MT-")
control3st3[["percent.mt"]] <- PercentageFeatureSet(object = control3st3, pattern = "^MT-")
p1 <- VlnPlot(object = control1st3, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3)
p2 <- VlnPlot(object = control2st3, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3)
p3 <- VlnPlot(object = control3st3, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3)
CombinePlots(plots = list(p1, p2, p3), ncol =1)
p4 <- FeatureScatter(object = control1st3, feature1 = "nCount_RNA", feature2 = "percent.mt")
p5 <- FeatureScatter(object = control1st3, feature1 = "nCount_RNA", feature2 = "nFeature_RNA")
p6 <- FeatureScatter(object = control2st3, feature1 = "nCount_RNA", feature2 = "percent.mt")
p7 <- FeatureScatter(object = control2st3, feature1 = "nCount_RNA", feature2 = "nFeature_RNA")
p8 <- FeatureScatter(object = control3st3, feature1 = "nCount_RNA", feature2 = "percent.mt")
p9 <- FeatureScatter(object = control3st3, feature1 = "nCount_RNA", feature2 = "nFeature_RNA")
CombinePlots(plots = list(p4, p5, p6, p7, p8, p9), ncol =2)
# Thresholding ####
# RNA based cell thresholding
# The operator can add columns to object metadata. This is a great place to stash QC stats
control1st3 <- subset(x = control1st3, subset = nFeature_RNA > 200 & nFeature_RNA < 8000 & percent.mt < 20)
control2st3 <- subset(x = control2st3, subset = nFeature_RNA > 200 & nFeature_RNA < 8000 & percent.mt < 20)
control3st3 <- subset(x = control3st3, subset = nFeature_RNA > 200 & nFeature_RNA < 8000 & percent.mt < 20)
p10 <- VlnPlot(object = control1st3, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3)
p11 <- VlnPlot(object = control2st3, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3)
p12 <- VlnPlot(object = control3st3, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3)
CombinePlots(plots = list(p10, p11, p12), ncol =1)
p13 <- FeatureScatter(object = control1st3, feature1 = "nCount_RNA", feature2 = "percent.mt")
p14 <- FeatureScatter(object = control1st3, feature1 = "nCount_RNA", feature2 = "nFeature_RNA")
p15 <- FeatureScatter(object = control2st3, feature1 = "nCount_RNA", feature2 = "percent.mt")
p16 <- FeatureScatter(object = control2st3, feature1 = "nCount_RNA", feature2 = "nFeature_RNA")
p17 <- FeatureScatter(object = control3st3, feature1 = "nCount_RNA", feature2 = "percent.mt")
p18 <- FeatureScatter(object = control3st3, feature1 = "nCount_RNA", feature2 = "nFeature_RNA")
CombinePlots(plots = list(p13, p14, p15, p16, p17, p18), ncol =2)
controlST3 <- sample(x = c("controlST3"), size = ncol(x = control1st3), replace = TRUE)
control1st3$groups <- controlST3
head(x = control1st3[[]])
controlST3 <- sample(x = c("controlST3"), size = ncol(x = control2st3), replace = TRUE)
control2st3$groups <- controlST3
head(x = control2st3[[]])
controlST3 <- sample(x = c("controlST3"), size = ncol(x = control3st3), replace = TRUE)
control3st3$groups <- controlST3
head(x = control3st3[[]])
# Merge Datasets
# Based on comment to Issue #4753 https://github.com/satijalab/seurat/issues/4753
# We use RPCA to yield conserved mapping
controldiabetes1 <- c(ctrl1st2, ctrl2st2, ctrl3st2, control1st3, control2st3, control3st3)
# normalize and identify variable features for each dataset independently
controldiabetes1.list <- lapply(X = controldiabetes1, FUN = function(x) {
x <- NormalizeData(x)
x <- FindVariableFeatures(x, selection.method = "vst", nfeatures = 3000)
})
# select features that are repeatedly variable across datasets for integration run PCA on each
# dataset using these features
features <- SelectIntegrationFeatures(object.list = controldiabetes1.list)
controldiabetes1.list <- lapply(X = controldiabetes1.list, FUN = function(x) {
x <- ScaleData(x, features = features, verbose = FALSE)
x <- RunPCA(x, features = features, verbose = FALSE)
})
# Perform integration
controldiabetes1.anchors <- FindIntegrationAnchors(object.list = controldiabetes1.list, anchor.features = features, reduction = "rpca")
controldiabetes1.combined <- IntegrateData(anchorset = controldiabetes1.anchors)
# Run the standard workflow for visualization and clustering
DefaultAssay(controldiabetes1.combined) <- "integrated"
controldiabetes1.combined <- ScaleData(controldiabetes1.combined, verbose = FALSE)
controldiabetes1.combined <- RunPCA(controldiabetes1.combined, npcs = 30, verbose = FALSE)
controldiabetes1.combined <- RunUMAP(controldiabetes1.combined, reduction = "pca", dims = 1:30)
controldiabetes1.combined <- FindNeighbors(controldiabetes1.combined, reduction = "pca", dims = 1:30)
controldiabetes1.combined <- FindClusters(controldiabetes1.combined, resolution = 0.5)
# Visualization
p1 <- DimPlot(controldiabetes1.combined, reduction = "umap")
p2 <- DimPlot(controldiabetes1.combined, reduction = "umap", label = TRUE,
repel = TRUE)
p1 + p2
# LINEAR DIMENSIONALITY REDUCTION ####
DefaultAssay(object = controldiabetes1.combined)
DefaultAssay(object = controldiabetes1.combined) <- "integrated"
controldiabetes1.combined <- RunPCA(object = controldiabetes1.combined, features = VariableFeatures(object = controldiabetes1.combined))
print(x = controldiabetes1.combined[["pca"]], dims = 1:5, nfeatures = 5)
VizDimLoadings(object = controldiabetes1.combined, dims = 1:2, reduction = "pca")
DimPlot(object = controldiabetes1.combined, reduction = "pca")
DimHeatmap(object = controldiabetes1.combined, dims = 1, cells = 500, balanced = TRUE)
DimHeatmap(object = controldiabetes1.combined, dims = 1:15, cells = 500, balanced = TRUE)
ElbowPlot(object = controldiabetes1.combined)
#CLUSTER ANALYSIS ####
# Clustering, we run multiple permutations to allow clustree to analyze optimal clustering resolution.
controldiabetes1.combined <- FindNeighbors(object = controldiabetes1.combined, dims = 1:20)
controldiabetes1.combined <- FindClusters(object = controldiabetes1.combined, resolution = 0)
controldiabetes1.combined <- FindClusters(object = controldiabetes1.combined, resolution = 0.1)
controldiabetes1.combined <- FindClusters(object = controldiabetes1.combined, resolution = 0.2)
controldiabetes1.combined <- FindClusters(object = controldiabetes1.combined, resolution = 0.3)
controldiabetes1.combined <- FindClusters(object = controldiabetes1.combined, resolution = 0.4)
controldiabetes1.combined <- FindClusters(object = controldiabetes1.combined, resolution = 0.5)
controldiabetes1.combined <- FindClusters(object = controldiabetes1.combined, resolution = 0.6)
# Cluster-tree analysis, looking appropriate non-anomalous clustering resolution
clustree(controldiabetes1.combined, prefix = "integrated_snn_res.")
controldiabetes1.combined <- FindClusters(object = controldiabetes1.combined, resolution = 0.4)
table(controldiabetes1.combined$seurat_clusters)
table(Idents(controldiabetes1.combined), controldiabetes1.combined$orig.ident)
controldiabetes1.combined <- RunUMAP(controldiabetes1.combined, dims = 1:30)
# DATA-VISUALIZATION ####
# note that you can set `label = TRUE` or use the LabelClusters function to help label
# individual clusters
p1010 <- DimPlot(object = controldiabetes1.combined, group.by = c("orig.ident", "integrated_snn_res.0.4"), combine = FALSE, pt.size = 1, raster = FALSE, label = TRUE)
p1010 <- lapply(X = p1010, FUN = function(x) x + theme(legend.position = "top") + guides(color = guide_legend(nrow = 3, byrow = TRUE, override.aes = list(size = 3))))
CombinePlots(p1010)
controldiabetes1u.combined <- RenameIdents(object= controldiabetes1.combined, `0` = "0", `1` = "1", `2` = "2",`3` = "3", `4` = "4", `5` = "4", `6` = "5", `7` = "6", `8` = "3", `9` = "3",`10` = "2", `11` = "7", `12` = "3", `13` = "8", `14` = "9", `15` = "10", `16` = "11", `17` = "12", `18` = "13")
DimPlot(controldiabetes1u.combined, label = TRUE, raster = FALSE)
#Renaming the clusters
controldiabetes.combined_newUMAP2 <- RenameIdents(controldiabetes1u.combined, `2` = "D1", `11` = "D2",
`1` = "D3", `10` = "D4",`9` = "D5", `3` = "A1", `5` = "A2", `0` = "A3", `4` = "A5",
`6` = "A6", `8` = "M", `12` = "I", `13` = "E", `7` = "S")
DimPlot(controldiabetes.combined_newUMAP2, label = TRUE)
DimPlot(controldiabetes.combined_newUMAP2, label = FALSE)
DimPlot(controldiabetes.combined_newUMAP2, reduction = "umap",
cols = c("turquoise1","mediumseagreen",
"royalblue1",
"lightslateblue",
"steelblue1",
"lightgoldenrod2",
"sienna3",
"yellow3",
"yellow",
"darkgoldenrod2",
"pink",
"red", "palegreen2", "grey"), label = TRUE, raster = FALSE,
pt.size = 1)
controldiabetes.combined_newUMAP2$CellType <- Idents(controldiabetes.combined_newUMAP2)
saveRDS(controldiabetes.combined_newUMAP2, file = "C:/Users/mad1188/Box/Mayur_shared_JDB/JUAN ONLY-DO NOT TOUCH/Mayur last/Materials and Methods/CONTROL_TimePoint_1+TimePoint_2.rds")
table((controldiabetes.combined_newUMAP2$CellType),(controldiabetes.combined_newUMAP2$orig.ident))
head(x = controldiabetes.combined_newUMAP2[[]])
write.csv([email protected], file='controldiabetes.combined_newUMAP2.csv', quote=T, row.names=T)
controldiabetes.combined_newUMAP24 <- subset(controldiabetes.combined_newUMAP2, idents = c("D1", "D2",
"D3", "D4","D5","A1","A2", "A3","A5",
"A6", "I"))
head(x = controldiabetes.combined_newUMAP24[[]])
controldiabetes1.combined.markers <- FindAllMarkers(controldiabetes1.combined, only.pos = TRUE, min.pct = 0.25, logfc.threshold = 0.41)
controldiabetes1.combined.markers %>%
group_by(clusters) %>%
slice_max(n = 2, order_by = avg_log2FC)
write.csv(controldiabetes1.combined.markers, 'C:/Users/mad1188/Box/Diabetes_new/All controls at Stage 2 and 3/Controlmarkers_celltype.csv')