-
Notifications
You must be signed in to change notification settings - Fork 0
/
Doublet_removed_Annotating_cells_from_merged_object.r
241 lines (163 loc) · 9.48 KB
/
Doublet_removed_Annotating_cells_from_merged_object.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
###############
# ====================================
# Author: Manu Singh
# Date: 2021-05-18
# Title: colon-seq data analysis for NIBD and CD patients
# ====================================
# =====================================
# Enviornment variables
# =====================================
setwd("~/Collab/Praveen/seurat_obj")
getwd()
Sys.Date()
main_dir <- getwd()
date <- gsub("-", "", Sys.Date())
dir.create(file.path(main_dir, date), showWarnings = FALSE)
setwd(file.path(main_dir, date))
getwd()
options(future.globals.maxSize = 6096*1524^2 )
set.seed(2811)
# =====================================
# Load libraries
# =====================================
library(Seurat)
library(ggplot2)
library(cowplot)
library(reticulate)
library(dplyr)
##############################
######## This is an initial clustering without normalizing for batch effect, which would be corrected later
#### The purpose of this clustering to get the initial annotations which we could later replaced by the original annotations of cell clusters
##### we are now dealing with the merged datasets .see line 96 of previous code
colon <- readRDS("Merged_seurat_objects_Praveen_NIBD_CD_samples")
# from the results of 95th quantiles
colon <- subset(colon, subset = nFeature_RNA > 3000 & nCount_RNA > 2.5e+04 & nCount_RNA < 1e+05)
# Removing the potential doublets
install.packages("remotes")
remotes::install_github("chris-mcginnis-ucsf/DoubletFinder")
library(DoubletFinder)
library(remotes)
pre_doublets <- colon
colon_normalized <- SCTransform(pre_doublets)
# parameters as per developer's recommendations
# results were consistent when we played from first 10 PCs till 30 PCs
sweep.res.list_colon <- paramSweep_v3(colon_normalized, PCs = 1:10, sct = FALSE)
sweep.stats_colon <- summarizeSweep(sweep.res.list_colon, GT = FALSE)
bcmvn_colon <- find.pK(sweep.stats_colon)
dev.off()
annotations <- [email protected]$orig.ident
homotypic.prop <- modelHomotypic(annotations)
nExp_poi <- round(0.075*nrow([email protected]))
nExp_poi.adj <- round(nExp_poi*(1-homotypic.prop))
colon_normalized <- doubletFinder_v3(colon_normalized, PCs = 1:10, pN = 0.25, pK = 0.09, nExp = nExp_poi.adj, reuse.pANN = "pANN_0.25_0.09_913", sct = FALSE)
Idents(colon_normalized) <- [email protected]$DF.classifications_0.25_0.09_899
length(WhichCells(colon_normalized, ident="Doublet"))
# less than 1% cells were doublets
jj <- WhichCells(colon_normalized, ident="Doublet")
doublets_removed <- pre_doublets[,!colnames(pre_doublets) %in% jj]
colon_normalized_doublet_removed <- SCTransform(doublets_removed)
colon <- colon_normalized_doublet_removed
### Normalizing in usual way
colon <- NormalizeData(colon, normalization.method = "LogNormalize", scale.factor = 20000)
# Here we took 3500 (combination of TEs and genes).
#In the case of genes or TEs alone, we only take 2000 rows as most variable
colon <- FindVariableFeatures(colon, selection.method = "vst", nfeatures = 3500)
# Here we are regressing out the impact of ribosomal, nUMI, and cell cycle genes
colon <- ScaleData(colon, vars.to.regress = c("S.Score", "G2M.Score", "nUMI", "ribo.genes"), features = row.names(colon))
colon <- RunPCA(colon, features = VariableFeatures(object = colon))
colon <- JackStraw(colon, num.replicate = 100)
colon <- ScoreJackStraw(colon, dims = 1:20)
colon <- FindNeighbors(colon, dims = 1:20)
colon <- FindClusters(colon, resolution = 0.8)
colon@misc$seurat_markers <- FindAllMarkers(colon, only.pos = TRUE, min.pct = 0.25, logfc.threshold = 0.25)
metadata <- as.data.frame(cbind(colnames(colon), as.character([email protected]$cell_type), as.character([email protected]$patients),[email protected]$nCount_RNA,[email protected]$nFeature_RNA))
colnames(metadata) <- c("coloumns", "cell_types", "Patients", "Total_RNA", "Total_features")
write.table(metadata, "MetaData_NIBD_CD_scRNAseq_for_Matt.tsv", row.names=F, col.names=T, sep="\t", dec=".", quote=F)
matt_is <- <- read.delim("~/Downloads/NIBD-CD_epithelial-cells_metadata.csv", stringsAsFactors=F, sep=",", row.names=1)
row.names(matt_is) <- gsub("_.$", "", row.names(matt_is))
head(row.names(matt_is))
#"NIBD_AAACCCAAGATCCCGC"
#"NIBD_AAACCCAAGCCTCCAG"
#"NIBD_AAACCCACAAGGCCTC"
#"NIBD_AAACCCACATGTTCAG"
#"NIBD_AAAGGATCACTTCATT"
#"NIBD_AAAGGATCAGCAGTCC"
unique(matt_is$cluster_named)
#"immature colonocyte"
#"CA1+ late colonocyte"
#"colonocyte progenitor"
#"CEACAM7+ colonocyte"
#"S phase TA"
#"immature goblet"
#"G2-M-G1 TA"
#"S phase TA-2"
#"goblet"
#"SPIB+ cells"
#"stem"
#"CA1+ early colonocyte"
#"Secretory progenitor"
#"EEC"
# Reannotating the cells based on the Epithelial cells classification used during the first submission of this paper
Idents(object = colon, cells = WhichCells(subset(colon, cells=row.names(subset(matt_is, cluster_named == "immature colonocyte"))))) <- "IC"
Idents(object = colon, cells = WhichCells(subset(colon, cells=row.names(subset(matt_is, cluster_named == "CA1+ late colonocyte"))))) <- "CA1+_late_colonocyte"
Idents(object = colon, cells = WhichCells(subset(colon, cells=row.names(subset(matt_is, cluster_named == "colonocyte progenitor"))))) <- "CP"
Idents(object = colon, cells = WhichCells(subset(colon, cells=row.names(subset(matt_is, cluster_named == "CEACAM7+ colonocyte"))))) <- "CEACAM7+_colonocyte"
Idents(object = colon, cells = WhichCells(subset(colon, cells=row.names(subset(matt_is, cluster_named == "S phase TA"))))) <- "S_phase_TA"
Idents(object = colon, cells = WhichCells(subset(colon, cells=row.names(subset(matt_is, cluster_named == "immature goblet"))))) <- "immature_goblet"
Idents(object = colon, cells = WhichCells(subset(colon, cells=row.names(subset(matt_is, cluster_named == "G2-M-G1 TA"))))) <- "G2_M_G1_TA"
Idents(object = colon, cells = WhichCells(subset(colon, cells=row.names(subset(matt_is, cluster_named == "S phase TA-2"))))) <- "S_phase_TA"
Idents(object = colon, cells = WhichCells(subset(colon, cells=row.names(subset(matt_is, cluster_named == "goblet"))))) <- "goblet"
Idents(object = colon, cells = WhichCells(subset(colon, cells=row.names(subset(matt_is, cluster_named == "SPIB+ cells"))))) <- "SPIB+_cells"
Idents(object = colon, cells = WhichCells(subset(colon, cells=row.names(subset(matt_is, cluster_named == "stem"))))) <- "Stem"
Idents(object = colon, cells = WhichCells(subset(colon, cells=row.names(subset(matt_is, cluster_named == "CA1+ early colonocyte"))))) <- "CA1+_early_colonocyte"
Idents(object = colon, cells = WhichCells(subset(colon, cells=row.names(subset(matt_is, cluster_named == "Secretory progenitor"))))) <- "secretory_progenitor"
Idents(object = colon, cells = WhichCells(subset(colon, cells=row.names(subset(matt_is, cluster_named == "EEC"))))) <- "EEC"
#Here, checking the status of seurat clusters after modifying the annotations.
DimPlot(colon)
dev.off()
#This visualization suggested that there are slightly higher number of cells in this analysis,
#so we are annotating the rest of cells carefully and only if they share the same clusters as classified above
Idents(object = colon, cells = WhichCells(colon, idents=18)) <- "Tuft_cells"
Idents(object = colon, cells = WhichCells(colon, idents=c(0,7))) <- "CA1+_late_colonocyte"
Idents(object = colon, cells = WhichCells(colon, idents=1)) <- "G2_M_G1_TA"
Idents(object = colon, cells = WhichCells(colon, idents=c(2,9))) <- "CEACAM7+_colonocyte"
Idents(object = colon, cells = WhichCells(colon, idents=3)) <- "secretory_progenitor"
Idents(object = colon, cells = WhichCells(colon, idents=4)) <- "S_phase_TA"
Idents(object = colon, cells = WhichCells(colon, idents=c(5,10))) <- "CA1+_early_colonocyte"
Idents(object = colon, cells = WhichCells(colon, idents=6)) <- "goblet"
Idents(object = colon, cells = WhichCells(colon, idents=8)) <- "immature_goblet"
Idents(object = colon, cells = WhichCells(colon, idents=11)) <- "SPIB+_cells"
Idents(object = colon, cells = WhichCells(colon, idents=12)) <- "CP"
Idents(object = colon, cells = WhichCells(colon, idents=13)) <- "IC"
Idents(object = colon, cells = WhichCells(colon, idents=14)) <- "Stem"
Idents(object = colon, cells = WhichCells(colon, idents=17)) <- "EEC"
# Note: cluster number 15 and 16 were immune cells or macrophages so were not annotated, and all the unannotated cells will be removed in next step
unique(Idents(colon))
#0 CA1+_late_colonocyte
#1 G2_M_G1_TA
#2 CEACAM7+_colonocyte
#3 secretory progenitor
#4 S_phase_TA, S_phase_TA-2
#5 CA1+_early_colonocyte
#6 goblet
#7 CA1+_late_colonocyte
#8 immature goblet
#9 CEACAM7+_colonocyte
#10 CA1+_early_colonocyte
#13 SPIB+_cells
#14 stem
#17 EEC
#### 18 Tuft cells
DefaultAssay(colon) <- "RNA"
colon <- subset(colon, idents = c("secretory_progenitor","goblet","CP","immature_goblet","CA1+_late_colonocyte","IC","G2_M_G1_TA","CA1+_late_colonocyte","SPIB+_cluster","Stem","S_phase_TA","Tuft_cells","CEACAM7+_colonocyte","EEC","CA1+_early_colonocyte"))
DimPlot(colon)
dev.off()
obj.list <- SplitObject(pbmc, split.by = "orig.ident")
### This is saving the individual samples, Note that these merged datasets were not corrected for batch effect
for(i in 1:length(obj.list)) {
colon_file <- obj.list[i]
colon_file_name <- basename(colon_file) %>% gsub(".rds","",.)
message(paste0("Processing Sample ",j," of ",length(Praveen_files)," - ", colon_file_name))
saveRDS(colon_file, file = paste0("Annotated_", colon_file_name, ".rds"))
}
saveRDS(colon, "Merged_Annotated_seurat_objects_Praveen_NIBD_CD_samples.rds")