Supplementary Figure 4

This notebook generates figures for Supplementary Figure 4

Preparation

You will need to download raw data from: https://lambrechtslab.sites.vib.be/en/immune-atlas

The page requires login to download data, but any Google account is good. After login, there will be multiple links for downloading. The 2069-Allcells.counts.rds is available in the “Clustering all cells” section. The 2076-OnlineTMEM106B.rds is in the “TMEM106B” section.

Data preparation

data <- readRDS("2076-OnlineTMEM106B.rds")
data@meta.data

obj <- readRDS("2069-Allcells.counts.rds")
meta.data <- data@meta.data[colnames(obj), ]
obj <- CreateSeuratObject(obj, meta.data = meta.data)
obj

An object of class Seurat 
33538 features across 65166 samples within 1 assay 
Active assay: RNA (33538 features, 0 variable features)

obj$Patient <- stringr::str_match(colnames(obj), "^(.+)_")[, 2]
table(obj$Patient, obj$PatientType)

        
           ICU  mild severe  Ward
  BAL001     0     0   1884     0
  BAL002     0  3775      0     0
  BAL003     0  2117      0     0
  BAL009     0  1212      0     0
  BAL010     0  6395      0     0
  BAL011     0   448      0     0
  BAL012     0     0    498     0
  BAL013     0     0   1950     0
  BAL014     0     0   1939     0
  BAL015     0     0   1989     0
  BAL016     0     0   1399     0
  BAL017     0  1128      0     0
  BAL018     0   504      0     0
  BAL019     0 11574      0     0
  BAL020     3     0   1405     0
  BAL021     0     0   2593     0
  BAL022    55     0   1427     0
  BAL023    71     0   1153     0
  BAL024     0     0   2267     0
  BAL025     0     0   1175     0
  BAL026   115     0   1996     0
  BAL027    40     0   1447     0
  BAL028     0  1183      0     0
  BAL029     0  1278      0     0
  BAL030     0   310      0     0
  BAL031     0     0   3795     0
  BAL032     0     0    582     0
  BAL033    31     0    974     0
  BAL034     0     0   1199     0
  BAL035     0     0   1144     0
  BAL036     0   230      0     0
  BAL037     0   896      0    18
  BAL038     0  1097      0     0
  BAL039     0     0   1017     0
  BAL040    10     0    843     0

obj$PatientType2 <- obj$PatientType
obj$PatientType2[obj$PatientType2 == 'ICU'] <- 'severe'
obj$PatientType2[obj$PatientType2 == 'Ward'] <- 'mild'

obj <- NormalizeData(obj, verbose = F)

Generate figures

library(Seurat)
library(cowplot)
library(grid)
library(gridExtra)
library(ggplot2)
library(lattice)


median.stat <- function(x){
   out <- quantile(x, probs = c(0.5))
   names(out) <- c("ymed")
   return(out) 
}

median.stat75 <- function(x){
   out <- quantile(x, probs = c(0.75))
   names(out) <- c("ymed")
   return(out) 
}

`%ni%`<- Negate(`%in%`)

library(ggpubr)

Warning: 程辑包‘ggpubr’是用R版本4.2.1 来建造的
载入程辑包：‘ggpubr’

The following object is masked from ‘package:cowplot’:

    get_legend

goi <- read.table("positive-candidates.txt")$V1

data <- obj[, obj$Domain == 'Epithelial']

data$Domain_disease_patient_type <- paste(data$Domain, data$Disease, data$PatientType2, sep='_')
data$Domain_disease_patient_type <- factor(data$Domain_disease_patient_type, 
                                          levels = c("Epithelial_control_mild", "Epithelial_control_severe",
                                                     "Epithelial_COVID19_mild", "Epithelial_COVID19_severe"
                                                     ))
Idents(data) <- 'Domain_disease_patient_type'

# goi <- readxl::read_xlsx("../../../data/Covid19/wauters-cell-research/2021-12-22 Candidate  list-COVID.XLSX", sheet = 2)
#cnt = 0
for (g in goi) {
  if (g %in% rownames(data)){
    png(paste0('positive/', g, '.png'), height = 6, width = 4, units = "in", res = 200)

    my_comparisons <- list( c("Epithelial_control_mild", "Epithelial_control_severe"), 
                        c("Epithelial_control_severe", "Epithelial_COVID19_mild"), 
                        c("Epithelial_COVID19_mild", "Epithelial_COVID19_severe"),
                        c("Epithelial_control_mild", "Epithelial_COVID19_mild"), 
                        c("Epithelial_control_severe", "Epithelial_COVID19_severe"),
                        c("Epithelial_control_mild", "Epithelial_COVID19_severe"))

    feature = g

    data2 <- data.frame(label = Idents(data), expression = data@assays$RNA@data[g, ])
    ymax = max(data2$expression)

    my_comparisons_y <- (c(1.5, 1.05, 1.15, 1.25, 1.4, 1.7) + 0.3) / 1.3
    print(
      ggplot(data=data2, aes(x = label, y=expression, fill = label)) + 
        geom_violin() +
        geom_point(position = position_jitter(seed = 1, width = 0.2), size = 0.1) + ylim(-0.01, ymax * 1.72) + 
        stat_summary(fun.y = mean, geom='point', size = 5, colour = "green", shape = 95) + 
        stat_summary(fun.y = median.stat75, geom='point', size = 5, colour = "red", shape = 95) + 
        stat_compare_means(comparisons = my_comparisons, label.y = ymax * my_comparisons_y, method = "wilcox.test", size = 4) +
        stat_compare_means(label.y = ymax * 1.7, label.x = 1.3, size = 4) +
        theme_classic() + theme(legend.position = "none") + ggtitle(g) + theme(axis.text.x = element_text(angle = 45, hjust = 1))
    )
    dev.off()
  } else {
    print(paste(g, "not found."))
  }
  #cnt = cnt + 1
  #if (cnt > 5) break
}

Warning: `fun.y` is deprecated. Use `fun` instead.Warning: `fun.y` is deprecated. Use `fun` instead.

library(ggpubr)

goi <- read.table("negative-candidates.txt")$V1

data <- obj[, obj$Domain == 'Epithelial']

data$Domain_disease_patient_type <- paste(data$Domain, data$Disease, data$PatientType2, sep='_')
data$Domain_disease_patient_type <- factor(data$Domain_disease_patient_type, 
                                          levels = c("Epithelial_control_mild", "Epithelial_control_severe",
                                                     "Epithelial_COVID19_mild", "Epithelial_COVID19_severe"
                                                     ))
Idents(data) <- 'Domain_disease_patient_type'

for (g in goi) {
  if (g %in% rownames(data)){
    png(paste0('negative/', g, '.png'), height = 6, width = 4, units = "in", res = 200)

    my_comparisons <- list( c("Epithelial_control_mild", "Epithelial_control_severe"), 
                        c("Epithelial_control_severe", "Epithelial_COVID19_mild"), 
                        c("Epithelial_COVID19_mild", "Epithelial_COVID19_severe"),
                        c("Epithelial_control_mild", "Epithelial_COVID19_mild"), 
                        c("Epithelial_control_severe", "Epithelial_COVID19_severe"),
                        c("Epithelial_control_mild", "Epithelial_COVID19_severe"))

    feature = g

    data2 <- data.frame(label = Idents(data), expression = data@assays$RNA@data[g, ])
    ymax = max(data2$expression)

    my_comparisons_y <- (c(1.5, 1.05, 1.15, 1.25, 1.4, 1.7) + 0.3) / 1.3
    print(
      ggplot(data=data2, aes(x = label, y=expression, fill = label)) + 
        geom_violin() +
        geom_point(position = position_jitter(seed = 1, width = 0.2), size = 0.1) + ylim(-0.01, ymax * 1.72) + 
        stat_summary(fun.y = mean, geom='point', size = 5, colour = "green", shape = 95) + 
        stat_summary(fun.y = median.stat75, geom='point', size = 5, colour = "red", shape = 95) + 
        stat_compare_means(comparisons = my_comparisons, label.y = ymax * my_comparisons_y, method = "wilcox.test", size = 4) +
        stat_compare_means(label.y = ymax * 1.7, label.x = 1.3, size = 4) +
        theme_classic() + theme(legend.position = "none") + ggtitle(g) + theme(axis.text.x = element_text(angle = 45, hjust = 1))
    )
    dev.off()
  } else {
    print(paste(g, "not found."))
  }
  #cnt = cnt + 1
  #if (cnt > 5) break
}

Warning: `fun.y` is deprecated. Use `fun` instead.Warning: `fun.y` is deprecated. Use `fun` instead.

[1] "RFWD2 not found."
[1] "CCDC101 not found."
[1] "WAPAL not found."
[1] "UFD1L not found."