title	author	date
Notebook2: all samples	Laura Wolbeck	2023-02-23

In the previous notebook the samples were split into early and late timepoints. Here we will generated the UMAP embedding for all samples together and provide the code for the respective figures.

Setup

#Load helper functions
source("scRNA_helper_functions_preterm.R")

#Load libraries
library(conos) #version 1.4.0
library(pagoda2)
library(magrittr)
library(ggplot2)
library(seurat)
library(cowplot)

# cms1 and cms2 are loaded from notebook1
cms <- append(cms1, cms2)

names = c("E18_5_1_",
          "E18_5_2_",
          "pre_P0_1",
          "pre_P0_2",
          "full_P2_1_",
          "full_P2_2_",
          "pre_P3_1",
          "pre_P3_2")

1. UMAP embedding and clustering

using conos to build the joint UMAP graph with forced alignment (alignment.strength = 0.3) to integrate samples well

con <- quickConos(cms,
                  names,
                  n.cores.p2=10,
                  n.cores.con=20, get.tsne = TRUE, alignment.strength=0.3)
con <- con$con

Additional clusters

Rerun leiden clustering to find additional clusters

con$findCommunities(method = leiden.community, resolution = 5, min.group.size = 15)
leiden_all <- con$clusters$leiden$groups %>% factor

The final clusters were generated using a combination of manual selection and findSubcommunities() function of conos to split the clusters further. The clusters were annotated using known cell-type specific marker genes from the literature.

2. Generate seurat object

generate seurat object to use seurats plotting functions

get raw counts from filtered cms

cm <- con$getJointCountMatrix(raw=T) %>% 
  Matrix::t() %>% 
  as.sparse()

Create Seurat object

seurat <- CreateSeuratObject(counts = cm, project = "All samples", min.cells = 0, min.features = 0)

add normalized cm

cm_norm <- con$getJointCountMatrix(raw=F) %>% 
  Matrix::t() %>% 
  as.sparse()

seurat <- SetAssayData(object= seurat, slot= "data", cm_norm )

to make sure that cells in annotations are in same order as in cm
read in anno and anno_rough

# order cells in "anno" as in "cm"
anno_rough %<>% .[match(colnames(cm), names(.))] 
anno %<>% .[match(colnames(cm), names(.))]

add annotations as metadata

seurat$annotation_rough <- anno_rough
seurat$annotation <- anno

create condition factor

sample <- con$getDatasetPerCell()

conditions <- gsub("full_P2_1_", "fullterm", sample)
conditions <- gsub("full_P2_2_", "fullterm", conditions)
conditions <- gsub("pre_P3_1", "preterm", conditions)
conditions <- gsub("pre_P3_2", "preterm", conditions)
conditions <- gsub("E18_5_1_", "fullterm", conditions)
conditions <- gsub("E18_5_2_", "fullterm", conditions)
conditions <- gsub("pre_P0_1", "preterm", conditions)
conditions <- gsub("pre_P0_2", "preterm", conditions)

conditions %<>% as.factor()

names(conditions) <- names(sample)

to make sure that cells in conditions are in same order as in cm

# order cells in "conditions" as in "cm"
conditions %<>% .[match(colnames(cm), names(.))]

add condition metadata to seurat object

seurat$condition <- conditions

3. Prepare data for adata object (python)

save cm and metadata to be read into python (for Fig S3B)

cm <- con$getJointCountMatrix(raw=F) %>% 
  as.sparse()
writeMM(cm, "cm_all.mtx")

genes <- colnames(cm)
write.csv(genes, 'all_genes.csv')

# order cells in "anno" as in "cm"
anno_rough %<>% .[match(rownames(cm), names(.))]

#prepare condition factor
sample <- con$getDatasetPerCell())
conditions <- gsub("full_P2_1_", "full_P2", sample)
conditions <- gsub("full_P2_2_", "full_P2", conditions)
conditions <- gsub("pre_P3_1", "pre_P3", conditions)
conditions <- gsub("pre_P3_2", "pre_P3", conditions)
conditions <- gsub("E18_5_1_", "E18_5", conditions)
conditions <- gsub("E18_5_2_", "E18_5", conditions)
conditions <- gsub("pre_P0_1", "pre_P0", conditions)
conditions <- gsub("pre_P0_2", "pre_P0", conditions)

# order cells in "conditions" as in "cm"
conditions %<>% .[match(colnames(cm), names(.))]

#save metadata in a dataframe
metadata <- data.frame(annotation=as.character(anno_rough), cellnames=names(anno_rough), sample=con$getDatasetPerCell(),condition=conditions)
head(metadata)
write.csv(metadata, 'all_metadata.csv')

Fig 1E

colours <- c("#FFAF32","#3232FF","#32FF32", "#FFFF32","#FF3232","#AF0F0F","#AF0F0F","#AF0F0F","#AF32FF","#E632E6","#FFC8FF","#DBFF00") %>% setNames(c("Immature_ependymal_cells", "RG","Intermediate_progenitors", "Dividing_cells", "Neuroblasts", "Neurons_1","Neurons_2","Neurons_3","OPC","Microglia","Endothelial/Pericytes/VSMC","Erythrocytes"))

con$plotGraph(groups=anno_rough, label="") + scale_colour_manual(values=colours)+  theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank())

Fig 1F

cowplot::plot_grid(con$plotGraph(gene="Glul", title="Glul", size=0.2) + theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(),panel.border=element_blank()), con$plotGraph(gene="Gls", title="Gls", size=0.2) + theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(),panel.border=element_blank()))

Fig 1G

colours <- c("#FF5D71", "#00B0F0", "#A9D18E", "#7030A0") %>% setNames(c("E18_5", "pre_P0", "full_P2", "pre_P3"))

level_order <- c("E18_5", "pre_P0", "full_P2", "pre_P3")
Idents(seurat) <- "condition" 
seurat$condition <- factor(x = seurat$condition, levels = level_order)

Idents(seurat) <- "annotation" 
VlnPlot(seurat, features="Glul",group.by = "condition", idents = "RG", cols = colours)  +
    stat_summary(fun.y = mean, geom='point', size = 3, colour = "yellow", show.legend = F)

VlnPlot(seurat, features="Gls",group.by = "condition", idents = "RG", cols = colours)  +
    stat_summary(fun.y = mean, geom='point', size = 3, colour = "yellow", show.legend = F)

Fig S3A

#defining colours of clusters
colours_high <- c("#e1af32","#FFAF32","#3232FF","#32FF32","#FFFF00","#CDCD32","#FF9b9b","#EB8C32","#C8961E","#FF3232","#AF3232","#B49191","#694646","#820000","#965050","#550505","#C80F0F","#AF32FF","#E632E6","#FFC8FF","#DBFF00") %>% setNames(c("Immature ependymal cells-2","Immature ependymal cells-2","RG","Intermediate progenitor cells","Dividing cells-2","Dividing cells-1","Neuroblasts-1", "Dividing cells-4", "Dividing cells-3","Neuroblasts-2","Neurons 1-1","Neurons 1-2","Neurons 2","Neurons 3-3","Neurons 3-4","Neurons 3-1","Neurons 3-2","OPC","Microglia", "Endothelial/Pericytes/VSMC","Erythrocytes"))

con$plotGraph(groups=anno, label="") + scale_colour_manual(values=colours_high)+  theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank())

Fig S3B and S3C

The following code is run in jupyter notebook (python)

import scanpy as sc
import pandas as pd
import numpy as np

# read in cm, genes and metadata
adata = sc.read_mtx("cm_all.mtx")
gene_df = pd.read_csv("all_genes.csv")
metadata = pd.read_csv("all_metadata.csv", index_col=0)

# preparing adata object
# adding gene names and cell names and metadata
adata.var_names = gene_df.x.values
adata.obs_names = metadata.cellnames.values
adata.obs = metadata

cluster_marker = ["Lrrc23", "Foxj1", "Slc1a3", "Fabp7", "Ascl1", "Mki67", "Cenpa", "Dcx","Rbfox3","Neurod6", "Nkx2-1","Ebf1","Drd2","Olig2","Mpeg1","Epas1", "Hbb-bs"]
sc.pl.dotplot(adata, cluster_marker , groupby='annotation', swap_axes=True, categories_order = ['Immature_ependymal_cells', 'RG' , 'Intermediate_progenitors','Dividing_cells', 'Neuroblasts','Neurons_1','Neurons_2','Neurons_3', 'OPC', 'Microglia',  'Endothelial/Pericytes/VSMC', 'Erythrocytes'], save="all.pdf")

sc.pl.dotplot(adata[adata.obs["condition"] == "E18_5"], cluster_marker ,title="E18.5", groupby='annotation', swap_axes=True, categories_order = ['Immature_ependymal_cells', 'RG' , 'Intermediate_progenitors','Dividing_cells', 'Neuroblasts','Neurons_1','Neurons_2','Neurons_3', 'OPC', 'Microglia',  'Endothelial/Pericytes/VSMC', 'Erythrocytes'], save="E18_5.pdf")

sc.pl.dotplot(adata[adata.obs["condition"] == "full_P2"], cluster_marker ,title="full_P2", groupby='annotation', swap_axes=True, categories_order = ['Immature_ependymal_cells', 'RG' , 'Intermediate_progenitors','Dividing_cells', 'Neuroblasts','Neurons_1','Neurons_2','Neurons_3', 'OPC', 'Microglia',  'Endothelial/Pericytes/VSMC', 'Erythrocytes'], save="full_P2.pdf")

sc.pl.dotplot(adata[adata.obs["condition"] == "pre_P0"], cluster_marker ,title="pre_P0", groupby='annotation', swap_axes=True, categories_order = ['Immature_ependymal_cells', 'RG' , 'Intermediate_progenitors','Dividing_cells', 'Neuroblasts','Neurons_1','Neurons_2','Neurons_3', 'OPC', 'Microglia',  'Endothelial/Pericytes/VSMC', 'Erythrocytes'], save="pre_P0.pdf")

sc.pl.dotplot(adata[adata.obs["condition"] == "pre_P3"], cluster_marker ,title="pre_P3", groupby='annotation', swap_axes=True, categories_order = ['Immature_ependymal_cells', 'RG' , 'Intermediate_progenitors','Dividing_cells', 'Neuroblasts','Neurons_1','Neurons_2','Neurons_3', 'OPC', 'Microglia',  'Endothelial/Pericytes/VSMC', 'Erythrocytes'], save="pre_P3.pdf")

Fig S3D

genes <- c("Foxj1", "Slc1a3","Ascl1", "Mki67", "Dcx", "Rbfox3", "Olig2", "Mpeg1", "Epas1")
genes %>% lapply(function(p) con$plotGraph(gene=p, title=p, size=0.2) + theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.border=element_blank())) %>%
  cowplot::plot_grid(plotlist=., ncol=5)

Fig S4

defining level orders

level_order_rough <- c("Immature_ependymal_cells", "RG","Intermediate_progenitors", "Dividing_cells","Neuroblasts", "Neurons_1", "Neurons_2", "Neurons_3" , "OPC" ,"Microglia", "Endothelial/Pericytes/VSMC", "Erythrocytes" )

level_order <- c("Immature ependymal cells-1", "Immature ependymal cells-2", "Radial glia","Intermediate progenitor cells", "Dividing cells-1", "Dividing cells-2","Dividing cells-3", "Dividing cells-4","Neuroblasts-1", "Neuroblasts-2","Neurons 1-1", "Neurons 1-2","Neurons 2", "Neurons 3-1" , "Neurons 3-2", "Neurons 3-3","Neurons 3-4",  "OPC" ,"Microglia", "Endothelial/Pericytes/VSMC", "Erythrocytes" )

#for A, C, E and G
Idents(seurat) <- "annotation_rough" 
seurat$annotation_rough <- factor(x = seurat$annotation_rough, levels = level_order_rough)

#for B, D, F and H
Idents(seurat) <- "annotation" 
seurat$annotation <- factor(x = seurat$annotation, levels = level_order)

Fig S4A and S4B

VlnPlot(seurat, features= "nFeature_RNA", pt.size = 0, log = F,) +  ggtitle(" ") + ylab("genes per cell") + xlab(" ")+ theme_bw(base_size = 16) + theme( legend.position = "none") +  theme(axis.text.x = element_text(angle = 45, hjust = 1)) + ylim(0, 12500)+
  theme(plot.margin = margin(0.5,0,0,1, "cm"))

Fig S4C and S4D

VlnPlot(seurat, features= "nCount_RNA", pt.size = 0)  + ggtitle(" ") +ylab("UMIs per cell") + xlab(" ") + theme_bw(base_size = 16) + theme(legend.position = "none") +  theme(axis.text.x = element_text(angle = 45, hjust = 1)) + 
  theme(plot.margin = margin(0.5,0,0,1, "cm")) + ylim(0, 100000)

Fig S4E

read in low resolution annotation as "anno_rough"

plotClusterBarplots(con, groups = anno_rough , show.entropy = F, show.size = F, sample.factor = conditions ) + theme(axis.text.x = element_text(angle = 45, hjust = 1)) + scale_fill_discrete(name = "condition") + 
  scale_x_discrete(limits = level_order_rough)+ 
  theme(plot.margin = margin(0.5,0,0,2, "cm"))

Fig S4F

read in high resolution annotation as "anno"

plotClusterBarplots(con, groups = anno , show.entropy = F, show.size = F, sample.factor = conditions) + theme(axis.text.x = element_text(angle = 45, hjust = 1)) + scale_fill_discrete(name = "condition") + 
  scale_x_discrete(limits = level_order)+ 
  theme(plot.margin = margin(0.5,0,0,2, "cm"))

Fig S4G

plotClusterBarplots(con, groups = anno_rough , show.entropy = F, show.size = F) + theme(axis.text.x = element_text(angle = 45, hjust = 1)) +  scale_x_discrete(limits = level_order_rough) + 
  theme(plot.margin = margin(0.5,0,0,2, "cm"))

Fig S4H

plotClusterBarplots(con, groups = anno, show.entropy = F, show.size = F, ) + theme(axis.text.x = element_text(angle = 45, hjust = 1)) + 
  scale_x_discrete(limits = level_order)+ 
  theme(plot.margin = margin(0.5,0,0,2, "cm"))

Fig S6

level_order <- c("E18_5", "pre_P0", "full_P2", "pre_P3")
Idents(seurat) <- "condition" 
seurat$condition <- factor(x = seurat$condition, levels = level_order)

Idents(seurat) <- "annotation"

function_vlnplot <- function(feature) {
  VlnPlot(seurat, features = feature, group.by = "condition", idents = "RG", cols = colours) +
    stat_summary(fun = mean, geom = "point", size = 3, colour = "yellow", show.legend = FALSE) +
    theme(axis.title.x = element_blank(),
          axis.title.y = element_blank(),
          legend.position = "none") +
    ggtitle(feature)
}

Fig S6A

features <- c( "Gapdh", "Ldha", "Eno1")

plots <- list()
for (x in features) {
  p <- function_vlnplot(x)
  plots[[x]] <- p
}

cowplot::plot_grid(plotlist = plots, ncol = 2)
ggsave("Vlnplots2_mean.pdf", width = 7, height = 8)

Fig S6B

VlnPlot(seurat, features="Glud1",group.by = "condition", idents = "RG", cols = colours)  +
    stat_summary(fun.y = mean, geom='point', size = 3, colour = "yellow", show.legend = F)

Fig S6C

features <- c("Mki67", "Sox2", "Gfap", "Nes", "Ascl1", "Dcx")

plots <- list()
for (x in features) {
  p <- function_vlnplot(x)
  plots[[x]] <- p
}

cowplot::plot_grid(plotlist = plots, ncol = 3)
ggsave("Vlnplots_mean.pdf")

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Notebook2_all_samples.md

Notebook2_all_samples.md

Setup

1. UMAP embedding and clustering

Additional clusters

2. Generate seurat object

3. Prepare data for adata object (python)

Fig 1E

Fig 1F

Fig 1G

Fig S3A

Fig S3B and S3C

Fig S3D

Fig S4

Fig S4A and S4B

Fig S4C and S4D

Fig S4E

Fig S4F

Fig S4G

Fig S4H

Fig S6

Fig S6A

Fig S6B

Fig S6C

Files

Notebook2_all_samples.md

Latest commit

History

Notebook2_all_samples.md

File metadata and controls

Setup

1. UMAP embedding and clustering

Additional clusters

2. Generate seurat object

3. Prepare data for adata object (python)

Fig 1E

Fig 1F

Fig 1G

Fig S3A

Fig S3B and S3C

Fig S3D

Fig S4

Fig S4A and S4B

Fig S4C and S4D

Fig S4E

Fig S4F

Fig S4G

Fig S4H

Fig S6

Fig S6A

Fig S6B

Fig S6C