手把手带你复现NC图表之Figure6
复现文章信息：
题目：Single-cell analysis reveals prognostic fibroblast subpopulations linked to molecular and immunological subtypes of lung cancer 期刊：Nature Communications 日期：2023年1月31日 DOI: 10.1038/s41467-023-35832-6
复现图——Figure 6
使用多个NSCLC队列进行生存分析
R包载入与数据准备
options(stringsAsFactors = F)
pardefault <- par()
library(Seurat)
library(tidyverse)
library(ggpubr)
library(ggsci)
library(survival)
library(survminer)
library(forestmodel)
data_directory <- "H:\\文献复现\\6\\"
source(paste0(data_directory, "0_NewFunctions.R"))
load(paste0(data_directory, "BulkData_Zenodo.Rdata"))
load(paste0(data_directory, "IntegratedFibs_Zenodo.Rdata"))
LUAD单变量COX回归
为了检测肌成纤维细胞丰度作为LUAD患者分层预后生物标志物的可能性，使用TCGA-LUAD数据集测试
#为了研究将肌成纤维细胞丰度作为LUAD患者分层的预后生物标志物的潜力，我们使用TCGA-LUAD数据集作为测试队列，以确定将样本分类为肌成纤维细胞高和低的最佳阈值
ys2test <- rev(seq(1,10, by = 1))
cox.zph_OK <- list()
for(DS in levels(Merged.LUADtraits$Dataset.factor)[]){
surv_data <- Merged.LUADtraits[Merged.LUADtraits$Sample.Subtype == "LUAD" &
Merged.LUADtraits$Dataset.factor == DS, ]
cox.zph_p = 0
for(i in ys2test){
if (cox.zph_p < 0.05) {
split_data <- survSplit(Surv(OS_YEARS, OS) ~
Myo_Fibs.pct,
zero = -0.1,
data = surv_data[!is.na(surv_data$OS_YEARS), ],
cut = i, episode = "tgroup", id = "id")
model.coxph2 <- coxph(Surv(OS_YEARS, OS) ~
Myo_Fibs.pct,
data = split_data[split_data$tgroup == 1,])
cox.zph_res = cox.zph(model.coxph2)
cox.zph_res.sum = c(i, cox.zph_res$table[1,3])
cox.zph_p <- cox.zph_res$table[1,3]
names(cox.zph_res.sum) <- c("OS_YEARS", "cox.zph_p")
}
cox.zph_OK[[DS]] <- cox.zph_res.sum
}
}
do.call(rbind, cox.zph_OK)
max_year <- min(do.call(rbind, cox.zph_OK)[,1])
#分别对每个数据集进行Cox回归分析
split_data <- list()
for(DS in levels(Merged.LUADtraits$Dataset.factor)[]){
surv_data <- Merged.LUADtraits[Merged.LUADtraits$Sample.Subtype == "LUAD" &
Merged.LUADtraits$Dataset.factor == DS, ]
split_data[[DS]] <- survSplit(Surv(OS_YEARS, OS) ~
Myo_Fibs.pct + Alveolar_Fibs.pct + Adventitial_Fibs.pct,
zero = -0.1,
data = surv_data[!is.na(surv_data$OS_YEARS), ],
cut = max_year, episode = "tgroup", id = "id")
}
Myo_cox.res <- list()
Myo_cox.res[["TCGA"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Myo_Fibs.pct),
data = split_data[["TCGA"]][split_data[["TCGA"]]$tgroup == 1,])
Myo_cox.res[["GSE68465"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Myo_Fibs.pct),
data = split_data[["GSE68465"]][split_data[["GSE68465"]]$tgroup == 1,])
Myo_cox.res[["GSE31210"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Myo_Fibs.pct),
data = split_data[["GSE31210"]][split_data[["GSE31210"]]$tgroup == 1,])
Myo_cox.res[["GSE72094"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Myo_Fibs.pct),
data = split_data[["GSE72094"]][split_data[["GSE72094"]]$tgroup == 1,])
forestmodel::forest_model(model_list = Myo_cox.res, panels = panels)
Myo_LUAD_Zscores <- list()
for(DS in names(Myo_cox.res)){
Myo_LUAD_Zscores[[DS]] <- summary(Myo_cox.res[[DS]])$coefficients[4:5]
}
Myo_LUAD_pvals <- list()
for(DS in names(Myo_cox.res)){
Myo_LUAD_pvals[[DS]] <- summary(Myo_cox.res[[DS]])$coefficients[5]
}
Myo_LUAD_pvals
metap::sumz(unlist(Myo_LUAD_pvals))
Alveolar_cox.res <- list()
Alveolar_cox.res[["TCGA"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Alveolar_Fibs.pct),
data = split_data[["TCGA"]][split_data[["TCGA"]]$tgroup == 1,])
Alveolar_cox.res[["GSE68465"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Alveolar_Fibs.pct),
data = split_data[["GSE68465"]][split_data[["GSE68465"]]$tgroup == 1,])
Alveolar_cox.res[["GSE31210"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Alveolar_Fibs.pct),
data = split_data[["GSE31210"]][split_data[["GSE31210"]]$tgroup == 1,])
Alveolar_cox.res[["GSE72094"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Alveolar_Fibs.pct),
data = split_data[["GSE72094"]][split_data[["GSE72094"]]$tgroup == 1,])
forestmodel::forest_model(model_list = Alveolar_cox.res, panels = panels)
Alveolar_LUAD_Zscores <- list()
for(DS in names(Alveolar_cox.res)){
Alveolar_LUAD_Zscores[[DS]] <- summary(Alveolar_cox.res[[DS]])$coefficients[4:5]
}
Alveolar_LUAD_pvals <- list()
for(DS in names(Alveolar_cox.res)){
Alveolar_LUAD_pvals[[DS]] <- summary(Alveolar_cox.res[[DS]])$coefficients[5]
}
Alveolar_LUAD_pvals
metap::sumz(unlist(Alveolar_LUAD_pvals))
Adventitial_cox.res <- list()
Adventitial_cox.res[["TCGA"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Adventitial_Fibs.pct),
data = split_data[["TCGA"]][split_data[["TCGA"]]$tgroup == 1,])
Adventitial_cox.res[["GSE68465"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Adventitial_Fibs.pct),
data = split_data[["GSE68465"]][split_data[["GSE68465"]]$tgroup == 1,])
Adventitial_cox.res[["GSE31210"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Adventitial_Fibs.pct),
data = split_data[["GSE31210"]][split_data[["GSE31210"]]$tgroup == 1,])
Adventitial_cox.res[["GSE72094"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Adventitial_Fibs.pct),
data = split_data[["GSE72094"]][split_data[["GSE72094"]]$tgroup == 1,])
forestmodel::forest_model(model_list = Adventitial_cox.res, panels = panels)
Adventitial_LUAD_Zscores <- list()
for(DS in names(Adventitial_cox.res)){
Adventitial_LUAD_Zscores[[DS]] <- summary(Adventitial_cox.res[[DS]])$coefficients[4:5]
}
LUAD_zScores.df <- data.frame(Z = do.call(rbind, Myo_LUAD_Zscores)[,1],
p = do.call(rbind, Myo_LUAD_Zscores)[,2],
SubPop = "Myo",
DS = rownames(do.call(rbind, Myo_LUAD_Zscores)))
LUAD_zScores.df <- rbind(LUAD_zScores.df,
data.frame(Z = do.call(rbind, Alveolar_LUAD_Zscores)[,1],
p = do.call(rbind, Alveolar_LUAD_Zscores)[,2],
SubPop = "Alveolar",
DS = rownames(do.call(rbind, Alveolar_LUAD_Zscores))))
LUAD_zScores.df <- rbind(LUAD_zScores.df,
data.frame(Z = do.call(rbind, Adventitial_LUAD_Zscores)[,1],
p = do.call(rbind, Adventitial_LUAD_Zscores)[,2],
SubPop = "Adventitial",
DS = rownames(do.call(rbind, Adventitial_LUAD_Zscores))))
LUAD_zScores.df$Subtype <- "LUAD"
reshape2::melt(LUAD_zScores.df, id.vars = c("SubPop", "DS"), measure.vars = "Z") %>%
ggplot(aes(y = value, x = SubPop, fill = SubPop)) +
geom_boxplot(outlier.shape = NA) +
geom_point(size = 1) +
theme_pubr(base_size = 7) +
scale_fill_manual(values = Fibs_col.palette) +
geom_hline(yintercept = 0, linetype = "dashed") +
ylab("Good Survival <- CoxPH Zscore -> Poor Survival") +
theme(axis.title.y = element_blank()) +
coord_flip()
LUSC单变量COX回归
ys2test <- rev(seq(1,10, by = 1))
cox.zph_OK <- list()
for(DS in unique(Merged.LUSCtraits$Dataset)){
surv_data <- Merged.LUSCtraits[Merged.LUSCtraits$Sample.Subtype == "LUSC" &
Merged.LUSCtraits$Dataset == DS, ]
cox.zph_p = 0
for(i in ys2test){
if (cox.zph_p < 0.05) {
split_data <- survSplit(Surv(OS_YEARS, OS) ~
Myo_Fibs.pct,# + Stage_4cat + Age,
zero = -0.1,
data = surv_data[!is.na(surv_data$OS_YEARS), ],
cut = i, episode = "tgroup", id = "id")
model.coxph2 <- coxph(Surv(OS_YEARS, OS) ~
Myo_Fibs.pct,
data = split_data[split_data$tgroup == 1,])
cox.zph_res = cox.zph(model.coxph2)
cox.zph_res.sum = c(i, cox.zph_res$table[1,3])
cox.zph_p <- cox.zph_res$table[1,3]
names(cox.zph_res.sum) <- c("OS_YEARS", "cox.zph_p")
}
cox.zph_OK[[DS]] <- cox.zph_res.sum
}
}
do.call(rbind, cox.zph_OK)
max_year <- 4
#分别对每个数据集进行Cox回归分析
split_data <- list()
for(DS in unique(Merged.LUSCtraits$Dataset)[]){
surv_data <- Merged.LUSCtraits[Merged.LUSCtraits$Sample.Subtype == "LUSC" &
Merged.LUSCtraits$Dataset == DS, ]
split_data[[DS]] <- survSplit(Surv(OS_YEARS, OS) ~
Myo_Fibs.pct + Alveolar_Fibs.pct + Adventitial_Fibs.pct,
zero = -0.1,
data = surv_data[!is.na(surv_data$OS_YEARS), ],
cut = max_year, episode = "tgroup", id = "id")
}
Myo_cox.res <- list()
Myo_cox.res[["TCGA"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Myo_Fibs.pct),
data = split_data[["TCGA"]][split_data[["TCGA"]]$tgroup == 1,])
Myo_cox.res[["GSE157009"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Myo_Fibs.pct),
data = split_data[["GSE157009"]][split_data[["GSE157009"]]$tgroup == 1,])
Myo_cox.res[["GSE157010"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Myo_Fibs.pct),
data = split_data[["GSE157010"]][split_data[["GSE157010"]]$tgroup == 1,])
Myo_cox.res[["GSE4573"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Myo_Fibs.pct),
data = split_data[["GSE4573"]][split_data[["GSE4573"]]$tgroup == 1,])
forestmodel::forest_model(model_list = Myo_cox.res, panels = panels)
Myo_LUSC_Zscores <- list()
for(DS in names(Myo_cox.res)){
Myo_LUSC_Zscores[[DS]] <- summary(Myo_cox.res[[DS]])$coefficients[4:5]
}
Alveolar_cox.res <- list()
Alveolar_cox.res[["TCGA"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Alveolar_Fibs.pct),
data = split_data[["TCGA"]][split_data[["TCGA"]]$tgroup == 1,])
Alveolar_cox.res[["GSE157009"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Alveolar_Fibs.pct),
data = split_data[["GSE157009"]][split_data[["GSE157009"]]$tgroup == 1,])
Alveolar_cox.res[["GSE157010"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Alveolar_Fibs.pct),
data = split_data[["GSE157010"]][split_data[["GSE157010"]]$tgroup == 1,])
Alveolar_cox.res[["GSE4573"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Alveolar_Fibs.pct),
data = split_data[["GSE4573"]][split_data[["GSE4573"]]$tgroup == 1,])
forestmodel::forest_model(model_list = Alveolar_cox.res, panels = panels)
Alveolar_LUSC_Zscores <- list()
for(DS in names(Alveolar_cox.res)){
Alveolar_LUSC_Zscores[[DS]] <- summary(Alveolar_cox.res[[DS]])$coefficients[4:5]
}
Adventitial_cox.res <- list()
Adventitial_cox.res[["TCGA"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Adventitial_Fibs.pct),
data = split_data[["TCGA"]][split_data[["TCGA"]]$tgroup == 1,])
Adventitial_cox.res[["GSE157009"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Adventitial_Fibs.pct),
data = split_data[["GSE157009"]][split_data[["GSE157009"]]$tgroup == 1,])
Adventitial_cox.res[["GSE157010"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Adventitial_Fibs.pct),
data = split_data[["GSE157010"]][split_data[["GSE157010"]]$tgroup == 1,])
Adventitial_cox.res[["GSE4573"]] <- coxph(Surv(OS_YEARS, OS) ~
scale(Adventitial_Fibs.pct),
data = split_data[["GSE4573"]][split_data[["GSE4573"]]$tgroup == 1,])
forestmodel::forest_model(model_list = Adventitial_cox.res, panels = panels)
Adventitial_LUSC_Zscores <- list()
for(DS in names(Adventitial_cox.res)){
Adventitial_LUSC_Zscores[[DS]] <- summary(Adventitial_cox.res[[DS]])$coefficients[4:5]
}
LUSC_zScores.df <- data.frame(Z = do.call(rbind, Myo_LUSC_Zscores)[,1],
p = do.call(rbind, Myo_LUSC_Zscores)[,2],
SubPop = "Myo",
DS = rownames(do.call(rbind, Myo_LUSC_Zscores)))
LUSC_zScores.df <- rbind(LUSC_zScores.df,
data.frame(Z = do.call(rbind, Alveolar_LUSC_Zscores)[,1],
p = do.call(rbind, Alveolar_LUSC_Zscores)[,2],
SubPop = "Alveolar",
DS = rownames(do.call(rbind, Alveolar_LUSC_Zscores))))
LUSC_zScores.df <- rbind(LUSC_zScores.df,
data.frame(Z = do.call(rbind, Adventitial_LUSC_Zscores)[,1],
p = do.call(rbind, Adventitial_LUSC_Zscores)[,2],
SubPop = "Adventitial",
DS = rownames(do.call(rbind, Adventitial_LUSC_Zscores))))
LUSC_zScores.df$Subtype <- "LUSC"
Figure 6A
散点图显示使用不同截点按肌成纤维细胞丰度对TCGA中LUAD25样本进行二分的标准化对数秩生存统计量（与总生存率的相关性）的变化。
Fig_6A <-
data.frame(stat = LUAD.Opt_cut$Myo_Fibs.pct$stats,
x = LUAD.Opt_cut$Myo_Fibs.pct$cuts) %>%
ggplot(aes(x = x, y = stat, colour = x < LUAD.Opt_cut$cutpoint[1,1])) +
geom_point(size = 2) +
theme_pubr(base_size = 15) +
theme(legend.position = "none") +
scale_colour_manual(values = c(pal_d3()(3)[3], "grey70")) +
xlab("Myo (% of all fibroblasts)") +
ylab("Standardized Log-Rank statistic") +
geom_vline(xintercept = LUAD.Opt_cut$cutpoint[1,1],
linetype = "dashed") +
annotate("label", y = 1, size = 5,
label = paste("Max ranked\nCutpoint =",
signif(LUAD.Opt_cut$cutpoint[1,1],3)),
x = LUAD.Opt_cut$cutpoint[1,1],
label.size = NA)
Fig_6A
Figure 6B
显示TCGA-LUAD中25样品中肌成纤维细胞丰度测量值分布的密度图
Fig_6B <-
Opt_cut.plot$Myo_Fibs.pct$distribution +
theme_pubr(base_size = 15) +
scale_fill_manual(values = c(pal_d3()(3)[3], "grey70"))+
scale_colour_manual(values = c(pal_d3()(3)[3], "grey70")) +
theme(legend.position = "none", plot.title = element_blank()) +
xlab("Myo (% of all fibroblasts)")
Fig_6B
Figure 6C
#生存曲线数据整理
Myo.ggsurvplot_list <- list()
for(i in levels(Merged.LUADtraits$Dataset.factor)){
cox_res <- summary(coxph(Surv(OS_YEARS, OS) ~ Myo_cat,
data = Merged.LUADtraits[Merged.LUADtraits$Sample.Subtype == "LUAD" &
Merged.LUADtraits$Dataset.factor == i, ]))
Myo.ggsurvplot_list[[i]] <- ggsurvplot(survfit(Surv(OS_YEARS, OS) ~ Myo_cat,
data = Merged.LUADtraits[Merged.LUADtraits$Sample.Subtype == "LUAD" &
Merged.LUADtraits$Dataset.factor == i, ]),
main = i,
pval = signif(cox_res$logtest[3],3), pval.coord = c(0,0.15), pval.size = 10,
conf.int = F,
risk.table = TRUE, risk.table.title = element_blank(),
risk.table.fontsize = 5,tables.height = 0.4,
risk.table.pos = "in",
font.tickslab = 17,
censor.shape = 124, censor.size = 4,
legend.labs = c("Low", "High"),
palette = c("grey70",pal_d3()(3)[3]),
ggtheme = theme_pubr(base_size = 17),
risk.table.y.col = T,
legend = c(0.85,0.85), legend.title = element_blank())
Myo.ggsurvplot_list[[i]]$plot <- Myo.ggsurvplot_list[[i]]$plot +
ggtitle(i) + ylab("Overall Survival Rate") +
theme(axis.text.x = element_blank(), axis.title.x = element_blank(),
plot.margin = margin(b= 0)) +
coord_cartesian(xlim = c(-0.3,NA))
Myo.ggsurvplot_list[[i]]$table <- Myo.ggsurvplot_list[[i]]$table + xlab("Time (Years)")+
theme(plot.margin = margin(t = 0), plot.background = element_blank()) +
coord_cartesian(xlim = c(-0.3,NA))
}
#Kaplan-Meier图显示TCGA-LUAD中25队列患者生存率，按肌成纤维细胞丰度使用二分最佳临界点分层
Fig_6C <-
ggarrange(Myo.ggsurvplot_list$TCGA$plot + theme(plot.title = element_blank()),
Myo.ggsurvplot_list$TCGA$table,
ncol = 1, nrow = 2, heights = c(0.7,0.3), align = "v",
common.legend = T, legend = "none")
Fig_6C
Figure 6D
#将该阈值应用于三个验证队列，证明了一致的显著患者分层
Fig_6D <-
ggarrange(Myo.ggsurvplot_list$GSE72094$plot, Myo.ggsurvplot_list$GSE31210$plot,
Myo.ggsurvplot_list$GSE68465$plot,
Myo.ggsurvplot_list$GSE72094$table, Myo.ggsurvplot_list$GSE31210$table,
Myo.ggsurvplot_list$GSE68465$table,
ncol = 3, nrow = 2, heights = c(0.7,0.3), align = "v",
common.legend = T, legend = "none")
Fig_6D
Figure 6E
散点图显示使用不同截点按肺泡成纤维细胞丰度对TCGA中LUAD25样本进行二分的标准化对数秩生存统计量（与总生存率的相关性）的变化。
Fig_6E <-
data.frame(stat = LUAD.Opt_cut$Alveolar_Fibs.pct$stats,
x = LUAD.Opt_cut$Alveolar_Fibs.pct$cuts) %>%
ggplot(aes(x = x, y = stat, colour = x < LUAD.Opt_cut$cutpoint[2,1])) +
geom_point(size = 2) +
theme_pubr(base_size = 15) +
theme(legend.position = "none") +
scale_colour_manual(values = c(pal_d3()(3)[1], "grey70")) +
xlab("Alveolar (% of all fibroblasts)") +
ylab("Standardized Log-Rank statistic") +
geom_vline(xintercept = LUAD.Opt_cut$cutpoint[2,1],
linetype = "dashed") +
annotate("label", y = 1, size = 5,
label = paste("Max ranked\nCutpoint =",
signif(LUAD.Opt_cut$cutpoint[2,1],3)),
x = LUAD.Opt_cut$cutpoint[2,1],
label.size = NA)
Fig_6E
Figure 6F
显示TCGA-LUAD中25样品中肺泡成纤维细胞丰度测量值分布的密度图
Fig_6F <-
Opt_cut.plot$Alveolar_Fibs.pct$distribution +
theme_pubr(base_size = 15) +
scale_fill_manual(values = c(pal_d3()(3)[1], "grey70"))+
scale_colour_manual(values = c(pal_d3()(3)[1], "grey70")) +
theme(legend.position = "none", plot.title = element_blank()) +
xlab("Alveolar (% of all fibroblasts)")
Fig_6F
Figure 6G
#生存曲线数据整理
Alv.ggsurvplot_list <- list()
for(i in levels(Merged.LUADtraits$Dataset.factor)){
cox_res <- summary(coxph(Surv(OS_YEARS, OS) ~ Alv_cat,
data = Merged.LUADtraits[Merged.LUADtraits$Sample.Subtype == "LUAD" &
Merged.LUADtraits$Dataset.factor == i, ]))
Alv.ggsurvplot_list[[i]] <- ggsurvplot(survfit(Surv(OS_YEARS, OS) ~ Alv_cat,
data = Merged.LUADtraits[Merged.LUADtraits$Sample.Subtype == "LUAD" &
Merged.LUADtraits$Dataset.factor == i, ]),
main = i,
pval = T, pval.coord = c(0,0.15), pval.size = 10,
conf.int = F,
risk.table = TRUE, risk.table.title = element_blank(),
risk.table.fontsize = 5,tables.height = 0.4,
risk.table.pos = "in",
font.tickslab = 17,
censor.shape = 124, censor.size = 4,
legend.labs = c("Low", "High"),
palette = c("grey70",pal_d3()(3)[1]),
ggtheme = theme_pubr(base_size = 17),
risk.table.y.col = T,
legend = c(0.85,0.85), legend.title = element_blank())
Alv.ggsurvplot_list[[i]]$plot <- Alv.ggsurvplot_list[[i]]$plot +
ggtitle(i) + ylab("Overall Survival Rate") +
theme(axis.text.x = element_blank(), axis.title.x = element_blank(),
plot.margin = margin(b= 0)) +
coord_cartesian(xlim = c(-0.3,NA))
Alv.ggsurvplot_list[[i]]$table <- Alv.ggsurvplot_list[[i]]$table + xlab("Time (Years)")+
theme(plot.margin = margin(t = 0), plot.background = element_blank()) +
coord_cartesian(xlim = c(-0.3,NA))
}
#Kaplan-Meier图显示TCGA-LUAD中25队列患者生存率，按肺泡成纤维细胞丰度使用二分最佳临界点分层
Fig_6G <-
ggarrange(Alv.ggsurvplot_list$TCGA$plot + theme(plot.title = element_blank()),
Alv.ggsurvplot_list$TCGA$table,
ncol = 1, nrow = 2, heights = c(0.7,0.3), align = "v",
common.legend = T, legend = "none")
Fig_6G
Figure 6H
将该阈值应用于三个验证队列，证明了一致的显著患者分层
Fig_6H <-
ggarrange(Alv.ggsurvplot_list$GSE72094$plot, Alv.ggsurvplot_list$GSE31210$plot,
Alv.ggsurvplot_list$GSE68465$plot,
Alv.ggsurvplot_list$GSE72094$table, Alv.ggsurvplot_list$GSE31210$table,
Alv.ggsurvplot_list$GSE68465$table,
ncol = 3, nrow = 2, heights = c(0.7,0.3), align = "v",
common.legend = T, legend = "none")
Fig_6H
多变量分类分析
森林图显示协变量独立风险比和上述分析的所有LUAD患者队列的4年总生存率多变量COX回归分析的校正p值，使用肌成纤维细胞丰度、疾病分期和患者年龄作为自变量
max_year <- 4
surv_data <- Merged.LUADtraits[Merged.LUADtraits$Dataset.factor %in% c("TCGA", "GSE68465", "GSE72094", "GSE31210"), ]
split_data <- survSplit(Surv(OS_YEARS, OS) ~
Myo_cat + Stage_4cat + Age,
zero = -0.1,
data = surv_data[!is.na(surv_data$OS_YEARS), ],
cut = max_year, episode = "tgroup", id = "id")
names(split_data)[1:2] <- c("Myo", "Stage")
Myo_cox.res <- coxph(Surv(OS_YEARS, OS) ~
Myo + Stage + Age,
data = split_data[split_data$tgroup == 1,])
summary(Myo_cox.res)$coefficients
Figure 6I
Fig_6I <- forest_model(Myo_cox.res,
format_options = forest_model_format_options(
banded = T, text_size = 5, point_size = 4))
Fig_6I
Figure 6J
#森林图显示协变量独立风险比（±95%置信区间）和上述所有LUAD患者队列4年总生存率的多变量考克斯回归分析的校正p值，使用肺泡成纤维细胞丰度、疾病分期和患者年龄作为自变量
surv_data <- Merged.LUADtraits[Merged.LUADtraits$Dataset.factor %in% c("TCGA", "GSE68465", "GSE72094", "GSE31210"), ]
split_data <- survSplit(Surv(OS_YEARS, OS) ~
Alv_cat + Stage_4cat + Age,
zero = -0.1,
data = surv_data[!is.na(surv_data$OS_YEARS), ],
cut = max_year, episode = "tgroup", id = "id")
names(split_data)[1:2] <- c("Alveolar", "Stage")
Alv_cox.res <- coxph(Surv(OS_YEARS, OS) ~
Alveolar + Stage + Age,
data = split_data[split_data$tgroup == 1,])
Fig_6J <- forest_model(Alv_cox.res,
format_options = forest_model_format_options(
banded = T, text_size = 2.5, point_size = 1))
Fig_6J
Figure 6
多个LUAD数据集中，肺泡和表皮成纤维细胞丰度与更好的总生存率相关。这种关联在肺泡成纤维细胞中尤其一致，在所有分析的数据集中都是显著的。因此，采用与上述相同的方法来测试将肺泡成纤维细胞丰度作为预后标志物的可能性。同样，这表明将LUAD队列分为肺泡成纤维细胞高或低在分层总生存率方面始终有效,并且这种关联与疾病分期和患者年龄无关。
本文参与 腾讯云自媒体同步曝光计划，分享自微信公众号。
原始发表：2023-06-16 22:12，如有侵权请联系 cloudcommunity@tencent.com 删除
评论
登录后参与评论
推荐阅读
目录
腾讯云开发者
