各种批量富集分析 cluster enrichments degs_enrich singlegroup_enrich clusterprolifer

#####安装archr包##别处复制
.libPaths(c("/home/data/t040413/R/x86_64-pc-linux-gnu-library/4.2",
            "/home/data/t040413/R/yll/usr/local/lib/R/site-library", 
            "/usr/local/lib/R/library",
            "/home/data/refdir/Rlib/"))
#BiocManager::install("Nebulosa")
#library("Nebulosa")
library("Seurat")
library(ggplot2)



getwd()

dir.create("/home/data/t040413/ipf/All_mesenchymal")
setwd("/home/data/t040413/ipf/All_mesenchymal")
getwd()



load("/home/data/t040413/ipf/gse157379_young_old_blm_mice/monocle/All.merge_final.rds")

table(All.merge$cell.type,All.merge$group)

DimPlot(All.merge,label = TRUE,repel = TRUE,label.size = 6)
# table(Idents(All.merge))
# All.merge$cell.type=Idents(All.merge)
# All.merge$cell.type=factor(All.merge$cell.type,levels = c(
#   
#   "Universal fibroblast",
#   "Specialized fibroblast",
#   "Interferon-responsive  fibroblast",
#   "Spp1 fibroblast",
#   "Mesenchymal progenitor",
#   "Hhip fibroblast",
#   "Pdgfrb fibroblast",
#   "Myofibroblast"
#   
# ))
# table(All.merge$cell.type)
# Idents(All.merge)=All.merge$cell.type
# save(All.merge,file = "/home/data/t040413/ipf/gse157379_young_old_blm_mice/monocle/All.merge_final.rds")


ggplot([email protected], 
       aes(x=stim, fill=Idents(All.merge))) + geom_bar(position = "fill")

ggplot([email protected], 
       aes(x=group, fill=Idents(All.merge))) + geom_bar(position = "fill")




specialized_degs=FindMarkers(All.merge,ident.1 ="Aged Fibrosis",ident.2 ="Aged",densify = TRUE,
                             subset.ident = "Specialized fibroblast",group.by = "group"   )

specialized_degs$gene=rownames(specialized_degs)

allmarkers_for_each=FindAllMarkers(All.merge,densify = TRUE )
allmarkers$gene=rownames(allmarkers)

openxlsx::write.xlsx(specialized_degs,file = "specialized_degs_agedfibrosis_vs_fibrosis.xlsx")

openxlsx::write.xlsx(allmarkers,file = "allmarkers_for_each.xlsx")




print(getwd())

VlnPlot(All.merge,features = "Inmt",split.by = "group",split.plot = TRUE)

VlnPlot(All.merge,features = c("Fn1","Inmt","Fap"),split.by = "group",split.plot = TRUE)

VlnPlot(All.merge,features = c("Acta2"),split.by = "group",split.plot = TRUE)

getwd()#"/home/data/t040413/ipf/All_mesenchymal/cluster_enrichments"
dir.create("./cluster_enrichments")
setwd("./cluster_enrichments/")



# 
# 
# allmarkers=allmarkers_for_each %>%
#   dplyr::group_by(cluster) %>%dplyr::filter(cluster=="Specialized fibroblast")

# 
# allmarkers_up=allmarkers[allmarkers$avg_log2FC>0,]
# allmarkers_up
# 
# 
# allmarkers_down=allmarkers[allmarkers$avg_log2FC<0,]
# allmarkers_down
# library(dplyr)
# special=specialized_degs %>%dplyr::filter(p_val_adj<0.05 &
#                                    ( avg_log2FC >0.5 | avg_log2FC <(-0.5)))
# 
# special_up=special[special$avg_log2FC>0,]
# special_down=special[special$avg_log2FC<0,]
# allmarkers
# 
# common=intersect(special$gene,allmarkers$gene)
# 
# common_up=intersect(special_up$gene,allmarkers$gene)
# 
# common_down=intersect(special_down$gene,allmarkers$gene)
# 
# 
# 
# alllist=list(common=common,common_down=common_down,common_up=common_up)
# 
# a_list=alllist
# names(a_list)
# 
# mergedlist=do.call(cbind, lapply(lapply(a_list, unlist), `length<-`, max(lengths(a_list))))
# #       cat  months dog 
# mergedlist=data.frame(mergedlist)
# openxlsx::write.xlsx(mergedlist,file = "meg.xlsx")
getwd()
allmarkers_for_each=openxlsx::read.xlsx("/home/data/t040413/ipf/All_mesenchymal/allmarkers_for_each.xlsx")
head(allmarkers_for_each)
table(allmarkers_for_each$cluster)
range(allmarkers_for_each$avg_log2FC)

specialized_degs=openxlsx::read.xlsx("/home/data/t040413/ipf/All_mesenchymal/specialized_degs_agedfibrosis_vs_fibrosis.xlsx")
head(specialized_degs)

setwd("../")
#################################################################1----------------------
###########单纯基因做富集分析single group enrichments----------
getwd()
dir.create("sing_groupgene_specialgene")
setwd("sing_groupgene_specialgene")
getwd()

genes_for_enrich_special_marker=
  allmarkers_for_each[allmarkers_for_each$cluster=="Specialized fibroblast",] %>%
dplyr::filter(avg_log2FC>0.5) %>% slice_max(avg_log2FC,n = 100) %>%   dplyr::select(gene) %>% .$'gene' %>% 
  gsub(pattern="\\.1",replacement = "",x = .,ignore.case = TRUE)


getwd() 
 a_list=list(genes_for_enrich_special_marker)
i=0
lapply(a_list,FUN = function(each_single_group)  {
  i=i+1
  group= paste(each_single_group[1:5],sep = "_",collapse = "_")
  group=paste(group,i)
  print(group)
  
  print(getwd())
  dir.create(paste0(group))
  setwd(paste0(group))
  print(getwd())
  
  #each_single_group=a_list[[1]]
  each_single_group=data.frame("gene"=each_single_group)
  
  
  library(clusterProfiler)
  library(org.Hs.eg.db) #人
  library(org.Mm.eg.db) #鼠
  library(ggplot2)
  # degs_for_nlung_vs_tlung$gene=rownames(degs_for_nlung_vs_tlung)
  # head(markers)
  # df=markers %>%dplyr::group_by(cluster)%>%
  #   filter(p_val_adj <0.05
  #   )
  # sce.markers=df
  # head(sce.markers)
  # print(getwd())
  
  sce.markers=each_single_group
  
  ids <- suppressWarnings(bitr(sce.markers[,1], 'SYMBOL', 'ENTREZID', 'org.Mm.eg.db'))
  head(ids)
  
  head(sce.markers)
  tail(sce.markers)
  dim(sce.markers)
  sce.markers=merge(sce.markers,ids,by.x='gene',by.y='SYMBOL')
  head(sce.markers)
  dim(sce.markers)
  # sce.markers$group=sce.markers$cluster
  # sce.markers=sce.markers[sce.markers$group!="none",]
  # dim(sce.markers)
  head(sce.markers)
  getwd()
  #    sce.markers=openxlsx::read.xlsx("/home/data/t040413/silicosis/fibroblast_myofibroblast/group_enrichments/")
  #sce.markers$cluster=sce.markers$mygroup
  dim(sce.markers)
  head(sce.markers)
  # gcSample=split(sce.markers$ENTREZID, sce.markers$cluster)
  
  gcSample=sce.markers$ENTREZI
  library(clusterProfiler)
  gcSample # entrez id , enrichGO
  print("===========开始go= All ont===========")
  xx <- enrichGO(gene = gcSample,OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                 pvalueCutoff=0.05) #organism="hsa",
  
  
  
  
  
  xx.BP <- enrichGO(gene = gcSample, ,OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                    pvalueCutoff=0.05,readable=TRUE,
                    ont="BP") #organism="hsa",
  p=clusterProfiler::dotplot(object = xx.BP,showCategory = 20,
                             label_format =60)
  p=p+ theme(axis.text.x = element_text(angle = 90,
                                        vjust = 0.5, hjust=0.5))
  p
  print(getwd())
  ggsave(paste0(group,'_degs_enrichGO-BP_enrichment--3.pdf'),plot = p,width = 14,height = 40,limitsize = F)
  ggsave(paste0(group,'_degs_enrichGO-BP_enrichment--3.png'),plot = p,width = 14,height = 40,limitsize = F)
  
  xx.CC <- enrichGO(gcSample ,OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                    pvalueCutoff=0.05,readable=TRUE,
                    ont="CC") #organism="hsa",
  p=clusterProfiler::dotplot(object = xx.CC,showCategory = 20,
                             label_format =60)
  p=p+ theme(axis.text.x = element_text(angle = 90,
                                        vjust = 0.5, hjust=0.5))
  p
  ggsave(paste0(group,'_degs_enrichGO-CC_enrichment--3.pdf'),plot = p,width = 14,height = 40,limitsize = F)
  ggsave(paste0(group,'_degs_enrichGO-CC_enrichment--3.png'),plot = p,width = 14,height = 40,limitsize = F)
  
  xx.MF <- enrichGO(gene = gcSample,OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                    pvalueCutoff=0.05,readable=TRUE,
                    ont="MF") #organism="hsa",
  p=clusterProfiler::dotplot(object = xx.MF,showCategory = 20,
                             label_format =60)
  p=p+ theme(axis.text.x = element_text(angle = 90,
                                        vjust = 0.5, hjust=0.5))
  p
  ggsave(paste0(group,'_degs_enrichGO-MF_enrichment--3.pdf'),plot = p,width = 14,height = 40,limitsize = F)
  ggsave(paste0(group,'_degs_enrichGO-MF_enrichment--3.png'),plot = p,width = 14,height = 40,limitsize = F)
  
  
  
  print(getwd())
  
  .libPaths()
  
  
  print("===========开始 kegg All ont============")
  gg<-clusterProfiler::enrichKEGG(gene = gcSample, #readable=TRUE,
                                  keyType = 'kegg',  #KEGG 富集
                                  organism='mmu',#"rno",
                                  pvalueCutoff = 0.05 #指定 p 值阈值（可指定 1 以输出全部
  )
  print("kegg---done------")
  #标准富集分析
  
  
  p=clusterProfiler::dotplot(object = xx,showCategory = 20,
                             label_format =100)
  p=p+ theme(axis.text.x = element_text(angle = 90,
                                        vjust = 0.5, hjust=0.5))
  p
  ggsave(paste0(group,'_degs_enrichGO-GO_enrichment--3.pdf'),plot = p,width = 14,height = 40,limitsize = F)
  ggsave(paste0(group,'_degs_enrichGO-GO_enrichment--3.png'),plot = p,width = 14,height = 40,limitsize = F)
  
  xx
  write.csv(xx,file = paste0(group,"enrichGO-GO_enrichment.csv"))
  
  p=clusterProfiler::dotplot(gg,showCategory = 20,
                             label_format = 40)
  p4=p+ theme(axis.text.x = element_text(angle = 90,
                                         vjust = 0.5, hjust=0.5))
  p4
  print(paste("保存位置",getwd(),sep = "  :   "))
  ggsave(paste0(group,'_degs_enrichGO-KEGG_enrichment-2.pdf'),plot = p4,width = 14,height = 25,limitsize = F)
  ggsave(paste0(group,'_degs_enrichGO-KEGG_enrichment-2.png'),plot = p4,width = 14,height = 25,limitsize = F)
  
  gg
  openxlsx::write.xlsx(gg,file = paste0(group,"_enrichGO-KEGG_enrichment.xlsx"))
  
  getwd()
  openxlsx::write.xlsx(sce.markers,file = paste0(group,"_sce.markers_for_each_clusterfor_enrichment.xlsx"))
  
  
  ##放大图片
  {
    getwd()
    #bp
    p=clusterProfiler::dotplot(object = xx.BP,showCategory = 100,
                               label_format =60)
    p=p+ theme(axis.text.x = element_text(angle = 90,
                                          vjust = 0.5, hjust=0.5))
    p
    ggsave(paste0(group,'_degs_enrichGO-BP_enrichment-100terms-3.pdf'),plot = p,width = 14,height = 100,limitsize = F)
    ggsave(paste0(group,'_degs_enrichGO-BP_enrichment-100terms-3.png'),plot = p,width = 14,height = 100,limitsize = F)
    print((getwd()))
    
    #cc
    p=clusterProfiler::dotplot(object = xx.CC,showCategory = 100,
                               label_format =60)
    p=p+ theme(axis.text.x = element_text(angle = 90,
                                          vjust = 0.5, hjust=0.5))
    p
    ggsave(paste0(group,'_degs_enrichGO-CC_enrichment-100terms-3.pdf'),plot = p,width = 14,height = 100,limitsize = F)
    ggsave(paste0(group,'_degs_enrichGO-CC_enrichment-100terms-3.png'),plot = p,width = 14,height = 100,limitsize = F)
    print((getwd()))
    
    #MF
    p=clusterProfiler::dotplot(object = xx.MF,showCategory = 100,
                               label_format =60)
    p=p+ theme(axis.text.x = element_text(angle = 90,
                                          vjust = 0.5, hjust=0.5))
    p
    ggsave(paste0(group,'_degs_enrichGO-MF_enrichment-100terms-3.pdf'),plot = p,width = 14,height = 100,limitsize = F)
    ggsave(paste0(group,'_degs_enrichGO-MF_enrichment-100terms-3.png'),plot = p,width = 14,height = 100,limitsize = F)
    print((getwd()))
    
    
    #KEGG
    p=clusterProfiler::dotplot(object = gg,showCategory = 100,
                               label_format =60)
    p=p+ theme(axis.text.x = element_text(angle = 90,
                                          vjust = 0.5, hjust=0.5))
    p
    ggsave(paste0(group,'_degs_enrichGO-KEGG_enrichment-100terms-3.pdf'),plot = p,width = 14,height = 100,limitsize = F)
    ggsave(paste0(group,'_degs_enrichGO-KEGG_enrichment-100terms-3.png'),plot = p,width = 14,height = 100,limitsize = F)
    print((getwd()))
    
    
    
  }
  
  #   save(xx.BP,xx.CC,xx.MF, gg, file = "~/silicosis/spatial_transcriptomicsharmony_cluster_0.5res_gsea/xx.Rdata")
  
  bp= xx.BP@result
  cc= xx.CC@result
  mf=xx.MF@result
  
  
  xx.all=do.call(rbind,list(bp,cc,mf))
  head(xx.all)
  openxlsx::write.xlsx(xx.all,file = paste0(group,"_enrichments_all.xlsx"))
  
  # save(xx.BP,xx.CC,xx.MF,xx.all,sce.markers,merged_degs, gg, file = "./xx.Rdata")
  
  result <- tryCatch({
    save(xx.BP,xx.CC,xx.MF,xx.all,sce.markers,merged_degs, gg, file = paste0(group,"_xx.Rdata"))
    # 在这里添加你希望在没有报错时执行的代码
    print("没有报错")
  }, error = function(e) {
    print(getwd())  # 在报错时执行的代码
    
    # 在这里添加你希望在报错时执行的额外代码
    save(xx.BP,xx.CC,xx.MF,xx.all,sce.markers, gg, file = paste0(group,"_xx_all.Rdata"))
  })
  
  # print(result)
  
  
  
  
  print(getwd())
  if (F) {
    #大鼠
    print("===========开始go============")
    xx <-clusterProfiler::enrichGO(gcSample, ,OrgDb="org.Rn.eg.db",
                                   readable=TRUE,
                                   ont = 'ALL',  #GO Ontology，可选 BP、MF、CC，也可以指定 ALL 同时计算 3 者
                                   pvalueCutoff=0.05) #organism="hsa", #'org.Hs.eg.db',
    
    print("===========开始 kegg============")
    gg<-clusterProfiler::enrichGO(gcSample,fun = "enrichKEGG",
                                  keyType = 'kegg',  #KEGG 富集
                                  organism="rno",
                                  pvalueCutoff = 0.05 #指定 p 值阈值（可指定 1 以输出全部
    )
    
    p=dotplot(xx)
    p2=p+ theme(axis.text.x = element_text(angle = 90,
                                           vjust = 0.5, hjust=0.5))
    p2
    ggsave('degs_enrichGO-GO_enrichment-2.pdf',plot = p2,width = 6,height = 20,limitsize = F)
    xx
    openxlsx::write.xlsx(xx,file = "enrichGO-GO_enrichment.xlsx")
    #
    p=dotplot(gg)
    p4=p+ theme(axis.text.x = element_text(angle = 90,
                                           vjust = 0.5, hjust=0.5))
    p4
    print(paste("保存位置",getwd(),sep = "  :   "))
    ggsave('degs_enrichGO-KEGG_enrichment-2.pdf',plot = p4,width = 6,height = 12,limitsize = F)
    gg
    openxlsx::write.xlsx(gg,file = "enrichGO-KEGG_enrichment.xlsx")
  }
  
  setwd("../")
})


enrich_Res=openxlsx::read.xlsx("./Inmt_Npnt_Mettl7a1_Macf1_Pcolce2 1/Inmt_Npnt_Mettl7a1_Macf1_Pcolce2 1_enrichments_all.xlsx")
head(enrich_Res)


degs_enrich=openxlsx::read.xlsx("./0.5/0.5_enrichments_all.xlsx")
head(degs_enrich)

##########gsea-----------

{
  #https://zhuanlan.zhihu.com/p/518144716
  #https://yulab-smu.top/biomedical-knowledge-mining-book/faq.html
  
  # library(org.Hs.eg.db)
  library(org.Mm.eg.db)
  library(clusterProfiler)
  library(enrichplot)
  library(tidyverse)
  library(ggstatsplot)
  {
    getwd()
    setwd("../")
    print(getwd())
    
    dir.create("cluster_special_genes_GSEA_kegg_go")
    setwd("cluster_special_genes_GSEA_kegg_go")
    
    ## 物种设置
    organism = 'mmu'    #  人类'hsa' 小鼠'mmu'   
    OrgDb = 'org.Mm.eg.db'#人类"org.Hs.eg.db" 小鼠"org.Mm.eg.db"
    
    head(allmarkers_for_each)
    marker_degs=allmarkers_for_each[allmarkers_for_each$cluster=="Specialized fibroblast",] %>%
      dplyr::filter(p_val_adj<0.05)  %>%
      dplyr::mutate(gene=
                      gsub(pattern="\\.1",replacement = "",x = .$gene))
   head(marker_degs)
     rownames(marker_degs)=marker_degs$gene
     marker_degs
    
    #### 按照需要可选择不同的DEG方法数据集 ####
    need_DEG <- marker_degs %>%
      dplyr::filter(p_val_adj<0.05)
    need_DEG <- need_DEG[,c(2,5)] #选择log2FoldChange和pvalue（凑成数据框）
    
    colnames(need_DEG) <- c('log2FoldChange','pvalue')
    need_DEG$SYMBOL <- rownames(need_DEG)
    
    ##### 创建gsea分析的geneList（包含从大到小排列的log2FoldChange和ENTREZID信息）####
    #转化id  
    df <- bitr(rownames(need_DEG), 
               fromType = "SYMBOL",
               toType =  "ENTREZID",
               OrgDb = OrgDb) #人数据库org.Hs.eg.db 小鼠org.Mm.eg.db
    need_DEG <- merge(need_DEG, df, by='SYMBOL')  #按照SYMBOL合并注释信息
    head(need_DEG) # https://yulab-smu.top/biomedical-knowledge-mining-book/faq.html
    geneList <- need_DEG$log2FoldChange
    names(geneList) <- need_DEG$ENTREZID
    geneList <- sort(geneList, decreasing = T)   #从大到小排序
    head(geneList)
    
    
    ##### gsea富集 ####
    KEGG_kk_entrez <- gseKEGG(geneList     = geneList,
                              organism     = organism, #人hsa 鼠mmu
                              pvalueCutoff = 0.25,
                              eps = 0)  #实际为padj阈值,可调整 
    KEGG_kk_clustergene <- DOSE::setReadable(KEGG_kk_entrez, 
                                 OrgDb=OrgDb,
                                 keyType='ENTREZID')#转化id             
    
    GO_kk_entrez <- gseGO(geneList     = geneList,
                          ont          = "ALL",  # "BP"、"MF"和"CC"或"ALL"
                          OrgDb        = OrgDb,#人类org.Hs.eg.db 鼠org.Mm.eg.db
                          keyType      = "ENTREZID",
                          pvalueCutoff = 0.25,
                          eps=0)   #实际为padj阈值可调整
    GO_kk_clustergene <- DOSE::setReadable(GO_kk_entrez, 
                               OrgDb=OrgDb,
                               keyType='ENTREZID')#转化id 
    print(getwd())
    save(KEGG_kk_entrez, GO_kk_entrez, 
         KEGG_kk_clustergene,GO_kk_clustergene,
         file = "cluster_special_gene_GSEA_result.RData")
    
    
    #首先对富集结果进行条件筛选，一般认为|NES|>1，NOM pvalue<0.05，FDR（padj）<0.25的通路是显著富集的；还可以从结果中细分出上下调通路单独绘图，以下代码仅展示KEGG通路富集结果的上调通路。 gseaplot2(）函数既可以对单独的通路绘图，也可以合并几个通路一起绘图；
    KEGG_kk_clustergene@result
    kegg_gsea_res=data.frame(KEGG_kk_clustergene@result)
    ##选取富集结果
    kk_gse <- KEGG_kk_clustergene
    kk_gse_entrez <- KEGG_kk_entrez
    
    ###条件筛选 
    #一般认为|NES|>1，NOM pvalue<0.05，FDR（padj）<0.25的通路是显著富集的
    kk_gse_cut <- kk_gse[kk_gse$pvalue<0.05 & kk_gse$p.adjust<0.25 & abs(kk_gse$NES)>1]
    kk_gse_cut_down <- kk_gse_cut[kk_gse_cut$NES < 0,]
    kk_gse_cut_up <- kk_gse_cut[kk_gse_cut$NES > 0,]
    
    #选择展现NES前几个通路 
    down_gsea <- kk_gse_cut_down[tail(order(kk_gse_cut_down$NES,decreasing = T),100),]
    up_gsea <- kk_gse_cut_up[head(order(kk_gse_cut_up$NES,decreasing = T),100),]
    diff_gsea <- kk_gse_cut[head(order(abs(kk_gse_cut$NES),decreasing = T),100),]
    
    
    #### 经典的GSEA图 uo----- 
    up_gsea$Description
    
    i=2
    gselist=list()
    for (i in 1:length(up_gsea$Description)) {
      
      gseap1 <- gseaplot2(kk_gse,
                          up_gsea$ID[i],#富集的ID编号
                          title = up_gsea$Description[i],#标题
                          color = "red", #GSEA线条颜色
                          base_size = 20,#基础字体大小
                          rel_heights = c(1.5, 0.5, 1),#副图的相对高度
                          subplots = 1:3,   #要显示哪些副图 如subplots=c(1,3) #只要第一和第三个图
                          ES_geom = "line", #enrichment score用线还是用点"dot"
                          pvalue_table = T) #显示pvalue等信息
      gselist[[up_gsea$Description[i]]]=gseap1 
      print(paste0(getwd(),up_gsea$Description[i]))
      
      pdf(file =  paste0(up_gsea$Description[i], 'cluster_special_gene_GSEA_up_1.pdf'), width =12, height =8)  
      print(gseap1)   
      dev.off()
      
    }
    
    
    pdf(file =  paste0("all_split", 'cluster_GSEA_up_1.pdf'), width =12, height =8)  
    for (i in names(gselist)) {
      print(gselist[[i]]) 
    }
    dev.off()
    
    ###下调gsea------
    {
      
      down_gsea$Description
      
      i=2
      gselist=list()
      for (i in 1:length(down_gsea$Description)) {
        
        gseap1 <- gseaplot2(kk_gse,
                            down_gsea$ID[i],#富集的ID编号
                            title = down_gsea$Description[i],#标题
                            color = "red", #GSEA线条颜色
                            base_size = 20,#基础字体大小
                            rel_heights = c(1.5, 0.5, 1),#副图的相对高度
                            subplots = 1:3,   #要显示哪些副图 如subplots=c(1,3) #只要第一和第三个图
                            ES_geom = "line", #enrichment score用线还是用点"dot"
                            pvalue_table = T) #显示pvalue等信息
        gselist[[down_gsea$Description[i]]]=gseap1 
        print(paste0(getwd(),down_gsea$Description[i]))
        
        pdf(file =  paste0(down_gsea$Description[i], '_cluster_GSEA_down_1.pdf'), width =12, height =8)  
        print(gseap1)   
        dev.off()
        
      }
      
      
      
      pdf(file =  paste0("all_split", '_cluster_GSEA_down_1.pdf'), width =12, height =8)  
      for (i in names(gselist)) {
        print(gselist[[i]]) 
      }
      dev.off()
      
    }
    
    #### 合并 GSEA通路 
    gseap2 <- gseaplot2(kk_gse,
                        up_gsea$ID,#富集的ID编号
                        title = "UP_GSEA_all",#标题
                        color = "red",#GSEA线条颜色
                        base_size = 20,#基础字体大小
                        rel_heights = c(1.5, 0.5, 1),#副图的相对高度
                        subplots = 1:3, #要显示哪些副图 如subplots=c(1,3) #只要第一和第三个图
                        ES_geom = "line",#enrichment score用线还是用点"dot"
                        pvalue_table = T) #显示pvalue等信息
    ggsave(gseap2, filename = "combined_GSEA_up_all.pdf",width =12,height =12)
    
    
    go_gsea_res=data.frame(GO_kk_clustergene@result)
    
  }
  
  
  
}



#########多个组放在list里面去做富集分析----------
lapply(a_list,head)

i=0
 lapply(a_list,FUN = function(each_single_group)  {
  i=i+1
  group= paste(each_single_group[1:5],sep = "_",collapse = "_")
  group=paste(group,i)
  print(group)

  print(getwd())
  dir.create(paste0(group))
  setwd(paste0(group))
  print(getwd())

  #each_single_group=a_list[[1]]
  each_single_group=data.frame("gene"=each_single_group)


  library(clusterProfiler)
  library(org.Hs.eg.db) #人
  library(org.Mm.eg.db) #鼠
  library(ggplot2)
  # degs_for_nlung_vs_tlung$gene=rownames(degs_for_nlung_vs_tlung)
  # head(markers)
  # df=markers %>%dplyr::group_by(cluster)%>%
  #   filter(p_val_adj <0.05
  #   )
  # sce.markers=df
  # head(sce.markers)
  # print(getwd())

  sce.markers=each_single_group

  ids <- suppressWarnings(bitr(sce.markers[,1], 'SYMBOL', 'ENTREZID', 'org.Mm.eg.db'))
  head(ids)

  head(sce.markers)
  tail(sce.markers)
  dim(sce.markers)
  sce.markers=merge(sce.markers,ids,by.x='gene',by.y='SYMBOL')
  head(sce.markers)
  dim(sce.markers)
  # sce.markers$group=sce.markers$cluster
  # sce.markers=sce.markers[sce.markers$group!="none",]
  # dim(sce.markers)
  head(sce.markers)
  getwd()
  #    sce.markers=openxlsx::read.xlsx("/home/data/t040413/silicosis/fibroblast_myofibroblast/group_enrichments/")
  #sce.markers$cluster=sce.markers$mygroup
  dim(sce.markers)
  head(sce.markers)
  # gcSample=split(sce.markers$ENTREZID, sce.markers$cluster)

  gcSample=sce.markers$ENTREZI
  library(clusterProfiler)
  gcSample # entrez id , enrichGO
  print("===========开始go= All ont===========")
  xx <- enrichGO(gene = gcSample,OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                 pvalueCutoff=0.05) #organism="hsa",





  xx.BP <- enrichGO(gene = gcSample, ,OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                    pvalueCutoff=0.05,readable=TRUE,
                    ont="BP") #organism="hsa",
  p=clusterProfiler::dotplot(object = xx.BP,showCategory = 20,
                             label_format =60)
  p=p+ theme(axis.text.x = element_text(angle = 90,
                                        vjust = 0.5, hjust=0.5))
  p
  print(getwd())
  ggsave(paste0(group,'_degs_enrichGO-BP_enrichment--3.pdf'),plot = p,width = 14,height = 40,limitsize = F)
  ggsave(paste0(group,'_degs_enrichGO-BP_enrichment--3.png'),plot = p,width = 14,height = 40,limitsize = F)

  xx.CC <- enrichGO(gcSample ,OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                    pvalueCutoff=0.05,readable=TRUE,
                    ont="CC") #organism="hsa",
  p=clusterProfiler::dotplot(object = xx.CC,showCategory = 20,
                             label_format =60)
  p=p+ theme(axis.text.x = element_text(angle = 90,
                                        vjust = 0.5, hjust=0.5))
  p
  ggsave(paste0(group,'_degs_enrichGO-CC_enrichment--3.pdf'),plot = p,width = 14,height = 40,limitsize = F)
  ggsave(paste0(group,'_degs_enrichGO-CC_enrichment--3.png'),plot = p,width = 14,height = 40,limitsize = F)

  xx.MF <- enrichGO(gene = gcSample,OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                    pvalueCutoff=0.05,readable=TRUE,
                    ont="MF") #organism="hsa",
  p=clusterProfiler::dotplot(object = xx.MF,showCategory = 20,
                             label_format =60)
  p=p+ theme(axis.text.x = element_text(angle = 90,
                                        vjust = 0.5, hjust=0.5))
  p
  ggsave(paste0(group,'_degs_enrichGO-MF_enrichment--3.pdf'),plot = p,width = 14,height = 40,limitsize = F)
  ggsave(paste0(group,'_degs_enrichGO-MF_enrichment--3.png'),plot = p,width = 14,height = 40,limitsize = F)



  print(getwd())

  .libPaths()


  print("===========开始 kegg All ont============")
  gg<-clusterProfiler::enrichKEGG(gene = gcSample, #readable=TRUE,
                                keyType = 'kegg',  #KEGG 富集
                                organism='mmu',#"rno",
                                pvalueCutoff = 0.05 #指定 p 值阈值（可指定 1 以输出全部
  )
  print("kegg---done------")
  #标准富集分析


  p=clusterProfiler::dotplot(object = xx,showCategory = 20,
                             label_format =100)
  p=p+ theme(axis.text.x = element_text(angle = 90,
                                        vjust = 0.5, hjust=0.5))
  p
  ggsave(paste0(group,'_degs_enrichGO-GO_enrichment--3.pdf'),plot = p,width = 14,height = 40,limitsize = F)
  ggsave(paste0(group,'_degs_enrichGO-GO_enrichment--3.png'),plot = p,width = 14,height = 40,limitsize = F)

  xx
  write.csv(xx,file = paste0(group,"enrichGO-GO_enrichment.csv"))

  p=clusterProfiler::dotplot(gg,showCategory = 20,
                             label_format = 40)
  p4=p+ theme(axis.text.x = element_text(angle = 90,
                                         vjust = 0.5, hjust=0.5))
  p4
  print(paste("保存位置",getwd(),sep = "  :   "))
  ggsave(paste0(group,'_degs_enrichGO-KEGG_enrichment-2.pdf'),plot = p4,width = 14,height = 25,limitsize = F)
  ggsave(paste0(group,'_degs_enrichGO-KEGG_enrichment-2.png'),plot = p4,width = 14,height = 25,limitsize = F)

  gg
  openxlsx::write.xlsx(gg,file = paste0(group,"_enrichGO-KEGG_enrichment.xlsx"))

  getwd()
  openxlsx::write.xlsx(sce.markers,file = paste0(group,"_sce.markers_for_each_clusterfor_enrichment.xlsx"))


  ##放大图片
  {
    getwd()
    #bp
    p=clusterProfiler::dotplot(object = xx.BP,showCategory = 100,
                               label_format =60)
    p=p+ theme(axis.text.x = element_text(angle = 90,
                                          vjust = 0.5, hjust=0.5))
    p
    ggsave(paste0(group,'_degs_enrichGO-BP_enrichment-100terms-3.pdf'),plot = p,width = 14,height = 100,limitsize = F)
    ggsave(paste0(group,'_degs_enrichGO-BP_enrichment-100terms-3.png'),plot = p,width = 14,height = 100,limitsize = F)
    print((getwd()))

    #cc
    p=clusterProfiler::dotplot(object = xx.CC,showCategory = 100,
                               label_format =60)
    p=p+ theme(axis.text.x = element_text(angle = 90,
                                          vjust = 0.5, hjust=0.5))
    p
    ggsave(paste0(group,'_degs_enrichGO-CC_enrichment-100terms-3.pdf'),plot = p,width = 14,height = 100,limitsize = F)
    ggsave(paste0(group,'_degs_enrichGO-CC_enrichment-100terms-3.png'),plot = p,width = 14,height = 100,limitsize = F)
    print((getwd()))

    #MF
    p=clusterProfiler::dotplot(object = xx.MF,showCategory = 100,
                               label_format =60)
    p=p+ theme(axis.text.x = element_text(angle = 90,
                                          vjust = 0.5, hjust=0.5))
    p
    ggsave(paste0(group,'_degs_enrichGO-MF_enrichment-100terms-3.pdf'),plot = p,width = 14,height = 100,limitsize = F)
    ggsave(paste0(group,'_degs_enrichGO-MF_enrichment-100terms-3.png'),plot = p,width = 14,height = 100,limitsize = F)
    print((getwd()))


    #KEGG
    p=clusterProfiler::dotplot(object = gg,showCategory = 100,
                               label_format =60)
    p=p+ theme(axis.text.x = element_text(angle = 90,
                                          vjust = 0.5, hjust=0.5))
    p
    ggsave(paste0(group,'_degs_enrichGO-KEGG_enrichment-100terms-3.pdf'),plot = p,width = 14,height = 100,limitsize = F)
    ggsave(paste0(group,'_degs_enrichGO-KEGG_enrichment-100terms-3.png'),plot = p,width = 14,height = 100,limitsize = F)
    print((getwd()))



  }

  #   save(xx.BP,xx.CC,xx.MF, gg, file = "~/silicosis/spatial_transcriptomicsharmony_cluster_0.5res_gsea/xx.Rdata")

 bp= xx.BP@result
 cc= xx.CC@result
  mf=xx.MF@result


  xx.all=do.call(rbind,list(bp,cc,mf))
  head(xx.all)
  openxlsx::write.xlsx(xx.all,file = paste0(group,"_enrichments_all.xlsx"))

  # save(xx.BP,xx.CC,xx.MF,xx.all,sce.markers,merged_degs, gg, file = "./xx.Rdata")

  result <- tryCatch({
    save(xx.BP,xx.CC,xx.MF,xx.all,sce.markers,merged_degs, gg, file = paste0(group,"_xx.Rdata"))
    # 在这里添加你希望在没有报错时执行的代码
    print("没有报错")
  }, error = function(e) {
    print(getwd())  # 在报错时执行的代码

    # 在这里添加你希望在报错时执行的额外代码
    save(xx.BP,xx.CC,xx.MF,xx.all,sce.markers, gg, file = paste0(group,"_xx_all.Rdata"))
  })

  # print(result)




  print(getwd())
  if (F) {
    #大鼠
    print("===========开始go============")
    xx <-clusterProfiler::enrichGO(gcSample, ,OrgDb="org.Rn.eg.db",
                                   readable=TRUE,
                                   ont = 'ALL',  #GO Ontology，可选 BP、MF、CC，也可以指定 ALL 同时计算 3 者
                                   pvalueCutoff=0.05) #organism="hsa", #'org.Hs.eg.db',

    print("===========开始 kegg============")
    gg<-clusterProfiler::enrichGO(gcSample,fun = "enrichKEGG",
                                  keyType = 'kegg',  #KEGG 富集
                                  organism="rno",
                                  pvalueCutoff = 0.05 #指定 p 值阈值（可指定 1 以输出全部
    )

    p=dotplot(xx)
    p2=p+ theme(axis.text.x = element_text(angle = 90,
                                           vjust = 0.5, hjust=0.5))
    p2
    ggsave('degs_enrichGO-GO_enrichment-2.pdf',plot = p2,width = 6,height = 20,limitsize = F)
    xx
    openxlsx::write.xlsx(xx,file = "enrichGO-GO_enrichment.xlsx")
    #
    p=dotplot(gg)
    p4=p+ theme(axis.text.x = element_text(angle = 90,
                                           vjust = 0.5, hjust=0.5))
    p4
    print(paste("保存位置",getwd(),sep = "  :   "))
    ggsave('degs_enrichGO-KEGG_enrichment-2.pdf',plot = p4,width = 6,height = 12,limitsize = F)
    gg
    openxlsx::write.xlsx(gg,file = "enrichGO-KEGG_enrichment.xlsx")
  }

  setwd("../")
})

 
 
 



##############################################################2-------------------------


{
  ##########################---------------------批量富集分析findallmarkers-enrichment analysis==================================================
  #https://mp.weixin.qq.com/s/WyT-7yKB9YKkZjjyraZdPg
  
  
  df=all_degs
  ##筛选阈值确定：p＜0.05，|log2FC|＞1
  p_val_adj = 0.05
  # avg_log2FC = 0.1
  
  fc=seq(0.8)
  print(getwd())
  #setwd("../")
  
  for (avg_log2FC in fc) {
    
    
    dir.create(paste0(avg_log2FC,"/"))
    setwd(paste0(avg_log2FC,"/"))
    
    
    print(getwd())
    print(paste0("Start----",avg_log2FC))
    
    head(all_degs)
    #根据阈值添加上下调分组标签：
    df$direction <- case_when(
      df$avg_log2FC > avg_log2FC & df$p_val_adj < p_val_adj ~ "up",
      df$avg_log2FC < -avg_log2FC & df$p_val_adj < p_val_adj ~ "down",
      TRUE ~ 'none'
    )
    head(df)
    
    df=df[df$direction!="none",]
    head(df)
    dim(df)
    df$mygroup=paste(df$group,df$direction,sep = "_")
    head(df)
    dim(df)
    ##########################----------------------enrichment analysis=
    #https://mp.weixin.qq.com/s/WyT-7yKB9YKkZjjyraZdPg
    {
      library(clusterProfiler)
      library(org.Hs.eg.db) #人
      library(org.Mm.eg.db) #鼠
      library(ggplot2)
      # degs_for_nlung_vs_tlung$gene=rownames(degs_for_nlung_vs_tlung)
      head(markers)
      df=markers %>%dplyr::group_by(cluster)%>%
        filter(p_val_adj <0.05
        )
      sce.markers=df
      head(sce.markers)
      print(getwd())
      ids <- suppressWarnings(bitr(sce.markers$gene, 'SYMBOL', 'ENTREZID', 'org.Mm.eg.db'))
      head(ids)
      
      head(sce.markers)
      tail(sce.markers)
      dim(sce.markers)
      sce.markers=merge(sce.markers,ids,by.x='gene',by.y='SYMBOL')
      head(sce.markers)
      dim(sce.markers)
      sce.markers$group=sce.markers$cluster
      sce.markers=sce.markers[sce.markers$group!="none",]
      dim(sce.markers)
      head(sce.markers)
      getwd()
      #    sce.markers=openxlsx::read.xlsx("/home/data/t040413/silicosis/fibroblast_myofibroblast/group_enrichments/")
      #sce.markers$cluster=sce.markers$mygroup
      dim(sce.markers)
      head(sce.markers)
      gcSample=split(sce.markers$ENTREZID, sce.markers$cluster)
      
      library(clusterProfiler)
      gcSample # entrez id , compareCluster 
      print("===========开始go= All ont===========")
      xx <- compareCluster(gcSample, fun="enrichGO",OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                           pvalueCutoff=0.05) #organism="hsa",
      
      
      xx.BP <- compareCluster(gcSample, fun="enrichGO",OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                              pvalueCutoff=0.05,readable=TRUE,
                              ont="BP") #organism="hsa",
      p=clusterProfiler::dotplot(object = xx.BP,showCategory = 20,
                                 label_format =60) 
      p=p+ theme(axis.text.x = element_text(angle = 90, 
                                            vjust = 0.5, hjust=0.5))
      p
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-BP_enrichment--3.pdf'),plot = p,width = 13,height = 40,limitsize = F)
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-BP_enrichment--3.png'),plot = p,width = 13,height = 40,limitsize = F)
      
      xx.CC <- compareCluster(gcSample, fun="enrichGO",OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                              pvalueCutoff=0.05,readable=TRUE,
                              ont="CC") #organism="hsa",
      p=clusterProfiler::dotplot(object = xx.CC,showCategory = 20,
                                 label_format =60) 
      p=p+ theme(axis.text.x = element_text(angle = 90, 
                                            vjust = 0.5, hjust=0.5))
      p
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-CC_enrichment--3.pdf'),plot = p,width = 13,height = 40,limitsize = F)
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-CC_enrichment--3.png'),plot = p,width = 13,height = 40,limitsize = F)
      
      xx.MF <- compareCluster(gcSample, fun="enrichGO",OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                              pvalueCutoff=0.05,readable=TRUE,
                              ont="MF") #organism="hsa",
      p=clusterProfiler::dotplot(object = xx.MF,showCategory = 20,
                                 label_format =60) 
      p=p+ theme(axis.text.x = element_text(angle = 90, 
                                            vjust = 0.5, hjust=0.5))
      p
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-MF_enrichment--3.pdf'),plot = p,width = 13,height = 40,limitsize = F)
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-MF_enrichment--3.png'),plot = p,width = 13,height = 40,limitsize = F)
      
      
      
      print(getwd())
      
      .libPaths()
      
      
      print("===========开始 kegg All ont============")
      gg<-clusterProfiler::compareCluster(gcSample,fun = "enrichKEGG",  #readable=TRUE,
                                          keyType = 'kegg',  #KEGG 富集
                                          organism='mmu',#"rno",
                                          pvalueCutoff = 0.05 #指定 p 值阈值（可指定 1 以输出全部
      )
      
      p=clusterProfiler::dotplot(object = xx,showCategory = 20,
                                 label_format =100) 
      p=p+ theme(axis.text.x = element_text(angle = 90, 
                                            vjust = 0.5, hjust=0.5))
      p
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-GO_enrichment--3.pdf'),plot = p,width = 13,height = 40,limitsize = F)
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-GO_enrichment--3.png'),plot = p,width = 13,height = 40,limitsize = F)
      
      xx
      write.csv(xx,file = paste0(avg_log2FC,"compareCluster-GO_enrichment.csv"))
      
      p=clusterProfiler::dotplot(gg,showCategory = 20,
                                 label_format = 40) 
      p4=p+ theme(axis.text.x = element_text(angle = 90, 
                                             vjust = 0.5, hjust=0.5))
      p4
      print(paste("保存位置",getwd(),sep = "  :   "))
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-KEGG_enrichment-2.pdf'),plot = p4,width = 13,height = 25,limitsize = F)
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-KEGG_enrichment-2.png'),plot = p4,width = 13,height = 25,limitsize = F)
      
      gg
      openxlsx::write.xlsx(gg,file = paste0(avg_log2FC,"_compareCluster-KEGG_enrichment.xlsx"))
      
      getwd()
      openxlsx::write.xlsx(sce.markers,file = paste0(avg_log2FC,"_sce.markers_for_each_clusterfor_enrichment.xlsx"))
      
      
      ##放大图片
      {
        getwd()
        #bp
        p=clusterProfiler::dotplot(object = xx.BP,showCategory = 100,
                                   label_format =60) 
        p=p+ theme(axis.text.x = element_text(angle = 90, 
                                              vjust = 0.5, hjust=0.5))
        p
        ggsave(paste0(avg_log2FC,'_degs_compareCluster-BP_enrichment-100terms-3.pdf'),plot = p,width = 13,height = 100,limitsize = F)
        ggsave(paste0(avg_log2FC,'_degs_compareCluster-BP_enrichment-100terms-3.png'),plot = p,width = 13,height = 100,limitsize = F)
        print((getwd()))
        
        #cc
        p=clusterProfiler::dotplot(object = xx.CC,showCategory = 100,
                                   label_format =60) 
        p=p+ theme(axis.text.x = element_text(angle = 90, 
                                              vjust = 0.5, hjust=0.5))
        p
        ggsave(paste0(avg_log2FC,'_degs_compareCluster-CC_enrichment-100terms-3.pdf'),plot = p,width = 13,height = 100,limitsize = F)
        ggsave(paste0(avg_log2FC,'_degs_compareCluster-CC_enrichment-100terms-3.png'),plot = p,width = 13,height = 100,limitsize = F)
        print((getwd()))
        
        #MF
        p=clusterProfiler::dotplot(object = xx.MF,showCategory = 100,
                                   label_format =60) 
        p=p+ theme(axis.text.x = element_text(angle = 90, 
                                              vjust = 0.5, hjust=0.5))
        p
        ggsave(paste0(avg_log2FC,'_degs_compareCluster-MF_enrichment-100terms-3.pdf'),plot = p,width = 13,height = 100,limitsize = F)
        ggsave(paste0(avg_log2FC,'_degs_compareCluster-MF_enrichment-100terms-3.png'),plot = p,width = 13,height = 100,limitsize = F)
        print((getwd()))
        
        
        #KEGG
        p=clusterProfiler::dotplot(object = gg,showCategory = 100,
                                   label_format =60) 
        p=p+ theme(axis.text.x = element_text(angle = 90, 
                                              vjust = 0.5, hjust=0.5))
        p
        ggsave(paste0(avg_log2FC,'_degs_compareCluster-KEGG_enrichment-100terms-3.pdf'),plot = p,width = 13,height = 100,limitsize = F)
        ggsave(paste0(avg_log2FC,'_degs_compareCluster-KEGG_enrichment-100terms-3.png'),plot = p,width = 13,height = 100,limitsize = F)
        print((getwd()))
        
        
        
      }
      
      #   save(xx.BP,xx.CC,xx.MF, gg, file = "~/silicosis/spatial_transcriptomicsharmony_cluster_0.5res_gsea/xx.Rdata")
      
      xx.BP@compareClusterResult$oncology="BP"
      xx.CC@compareClusterResult$oncology="CC"
      xx.MF@compareClusterResult$oncology="MF"
      
      
      xx.all=do.call(rbind,list(xx.BP@compareClusterResult,
                                xx.CC@compareClusterResult,
                                xx.MF@compareClusterResult))
      head(xx.all)    
      openxlsx::write.xlsx(xx.all,file = paste0(avg_log2FC,"_enrichments_all.xlsx"))    
      
      # save(xx.BP,xx.CC,xx.MF,xx.all,sce.markers,merged_degs, gg, file = "./xx.Rdata")
      
      result <- tryCatch({
        save(xx.BP,xx.CC,xx.MF,xx.all,sce.markers,merged_degs, gg, file = paste0(avg_log2FC,"_xx.Rdata"))
        # 在这里添加你希望在没有报错时执行的代码
        print("没有报错")
      }, error = function(e) {
        print(getwd())  # 在报错时执行的代码
        
        # 在这里添加你希望在报错时执行的额外代码
        save(xx.BP,xx.CC,xx.MF,xx.all,sce.markers, gg, file = paste0(avg_log2FC,"_xx_all.Rdata"))
      })
      
      # print(result)
      
      
      
      
      print(getwd())
      if (F) { 
        #大鼠
        print("===========开始go============")
        xx <-clusterProfiler::compareCluster(gcSample, fun="enrichGO",OrgDb="org.Rn.eg.db",
                                             readable=TRUE,
                                             ont = 'ALL',  #GO Ontology，可选 BP、MF、CC，也可以指定 ALL 同时计算 3 者
                                             pvalueCutoff=0.05) #organism="hsa", #'org.Hs.eg.db',
        
        print("===========开始 kegg============")
        gg<-clusterProfiler::compareCluster(gcSample,fun = "enrichKEGG",
                                            keyType = 'kegg',  #KEGG 富集
                                            organism="rno",
                                            pvalueCutoff = 0.05 #指定 p 值阈值（可指定 1 以输出全部
        )
        
        p=dotplot(xx) 
        p2=p+ theme(axis.text.x = element_text(angle = 90, 
                                               vjust = 0.5, hjust=0.5))
        p2
        ggsave('degs_compareCluster-GO_enrichment-2.pdf',plot = p2,width = 6,height = 20,limitsize = F)
        xx
        openxlsx::write.xlsx(xx,file = "compareCluster-GO_enrichment.xlsx")
        #
        p=dotplot(gg) 
        p4=p+ theme(axis.text.x = element_text(angle = 90, 
                                               vjust = 0.5, hjust=0.5))
        p4
        print(paste("保存位置",getwd(),sep = "  :   "))
        ggsave('degs_compareCluster-KEGG_enrichment-2.pdf',plot = p4,width = 6,height = 12,limitsize = F)
        gg
        openxlsx::write.xlsx(gg,file = "compareCluster-KEGG_enrichment.xlsx")   
      }  
      
      setwd("../")
    }
    
  }
  
  
  
  
}




###############################################################################3----------


##findallmarkers做富集分析 每个cluster的不同组别之间---------------


load("/home/data/t040413/ipf/gse157379_young_old_blm_mice/monocle/All.merge_final.rds")

All.merge$cell_type=All.merge$cell.type
subset_data=All.merge

fibroblast=subset_data
head([email protected])
table(fibroblast$group)

degs=list()
#Idents(fibroblast)=fibroblast$group
for (cluster in unique(fibroblast$cell_type)  ) {
  
  Idents(fibroblast)=fibroblast$cell_type
  subset_data=base::subset(fibroblast,idents = cluster)
  
  
  for (eachgroup in c('Aged Fibrosis' )) {
    
    Idents(subset_data)=subset_data$group
    
    
    try({
      degs[[paste0(cluster,eachgroup)]] = FindMarkers(subset_data,ident.1 = eachgroup,densify = TRUE,
                                                      ident.2 = "Aged"
      )
      
      degs[[paste0(cluster,eachgroup)]]$group=paste0(cluster,"_",eachgroup,"_vs_NS")
      
    }, silent = TRUE)  # 设置 silent = TRUE 以捕获错误但不中断执行
    
    
    
    print(paste0(cluster," __",eachgroup,"======done!========="))
  }
  
  
  print(paste0(" __",cluster,"======done!========="))
  # degs[[eachgroup]]$group=paste0(eachgroup,"_VS_NS")
  print(getwd())
}



head(degs)
names(degs)

degs=lapply(degs, function(x){
  x$gene=rownames(x)
  return(x)
})
head(degs)


merged_degs=do.call(rbind,degs)
head(merged_degs)

markers=merged_degs

############################针对性 修改sepcialized degs-------------------
markers=specialized_degs

print(getwd())
setwd("../")
print(getwd())
head(markers)

dir.create("./special_degs_for_fibrosis_vs_ns")
setwd("./special_degs_for_fibrosis_vs_ns")
print(getwd())



{
  print(getwd())
  
  all_degs=markers
  df=all_degs
  ##筛选阈值确定：p＜0.05，|log2FC|＞1
  p_val_adj = 0.05
  #  avg_log2FC = 0.8
  head(all_degs)
  
  various_fc= seq(0.5,0.8)
  
  
  for (avg_log2FC in various_fc) {
    
    
    dir.create(paste0(avg_log2FC,"/"))
    setwd(paste0(avg_log2FC,"/"))
    
    
    
    
    
    #根据阈值添加上下调分组标签：
    df$direction <- case_when(
      df$avg_log2FC > avg_log2FC & df$p_val_adj < p_val_adj ~ "up",
      df$avg_log2FC < -avg_log2FC & df$p_val_adj < p_val_adj ~ "down",
      TRUE ~ 'none'
    )
    head(df)
    print(getwd())
    df=df[df$direction!="none",]
    head(df)
    dim(df)
    df$mygroup=paste(df$group,df$direction,sep = "_")
    head(df)
    table(df$mygroup)
    dim(df)
    
    print(getwd())
    
    # dir.create("./degs_enrichments1fc")
    # setwd("./degs_enrichments1fc")
    ##########################----------------------enrichment analysis==
    #https://mp.weixin.qq.com/s/WyT-7yKB9YKkZjjyraZdPg
    
    
    {
      library(clusterProfiler)
      library(org.Hs.eg.db) #人
      library(org.Mm.eg.db) #鼠
      library(ggplot2)
      # degs_for_nlung_vs_tlung$gene=rownames(degs_for_nlung_vs_tlung)
      head(df)
      df$cluster=df$mygroup
      
      df=df %>%dplyr::group_by(cluster)%>%
        filter(p_val_adj <0.05)
      sce.markers=df
      head(sce.markers)
      print(getwd())
      ids <- suppressWarnings(bitr(sce.markers$gene, 'SYMBOL', 'ENTREZID', 'org.Mm.eg.db'))
      head(ids)
      
      head(sce.markers)
      tail(sce.markers)
      dim(sce.markers)
      sce.markers=merge(sce.markers,ids,by.x='gene',by.y='SYMBOL')
      head(sce.markers)
      dim(sce.markers)
      sce.markers$group=sce.markers$cluster
      sce.markers=sce.markers[sce.markers$group!="none",]
      dim(sce.markers)
      head(sce.markers)
      #sce.markers=openxlsx::read.xlsx("~/silicosis/spatial_transcriptomicsharmony_cluster_0.5res_gsea/sce.markers_for_each_clusterfor_enrichment.xlsx")
      #sce.markers$cluster=sce.markers$mygroup
      dim(sce.markers)
      head(sce.markers)
      gcSample=split(sce.markers$ENTREZID, sce.markers$cluster)
      
      library(clusterProfiler)
      library(org.Mm.eg.db)
      gcSample # entrez id , compareCluster 
      names(gcSample)
      print("===========开始go============")
      print("===========开始go= All ont===========")
      xx <- compareCluster(gcSample, fun="enrichGO",OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                           pvalueCutoff=0.05) #organism="hsa",
      
      
      xx.BP <- compareCluster(gcSample, fun="enrichGO",OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                              pvalueCutoff=0.05,readable=TRUE,
                              ont="BP") #organism="hsa",
      p=clusterProfiler::dotplot(object = xx.BP,showCategory = 20,
                                 label_format =60) 
      p=p+ theme(axis.text.x = element_text(angle = 90, 
                                            vjust = 0.5, hjust=0.5))
      p
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-BP_enrichment--3.pdf'),plot = p,width = 13,height = 40,limitsize = F)
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-BP_enrichment--3.png'),plot = p,width = 13,height = 40,limitsize = F)
      
      xx.CC <- compareCluster(gcSample, fun="enrichGO",OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                              pvalueCutoff=0.05,readable=TRUE,
                              ont="CC") #organism="hsa",
      p=clusterProfiler::dotplot(object = xx.CC,showCategory = 20,
                                 label_format =60) 
      p=p+ theme(axis.text.x = element_text(angle = 90, 
                                            vjust = 0.5, hjust=0.5))
      p
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-CC_enrichment--3.pdf'),plot = p,width = 13,height = 40,limitsize = F)
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-CC_enrichment--3.png'),plot = p,width = 13,height = 40,limitsize = F)
      
      xx.MF <- compareCluster(gcSample, fun="enrichGO",OrgDb="org.Mm.eg.db" ,   #'org.Hs.eg.db',
                              pvalueCutoff=0.05,readable=TRUE,
                              ont="MF") #organism="hsa",
      p=clusterProfiler::dotplot(object = xx.MF,showCategory = 20,
                                 label_format =60) 
      p=p+ theme(axis.text.x = element_text(angle = 90, 
                                            vjust = 0.5, hjust=0.5))
      p
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-MF_enrichment--3.pdf'),plot = p,width = 13,height = 40,limitsize = F)
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-MF_enrichment--3.png'),plot = p,width = 13,height = 40,limitsize = F)
      
      
      
      print(getwd())
      
      .libPaths()
      
      
      print("===========开始 kegg All ont============")
      gg<-clusterProfiler::compareCluster(gcSample,fun = "enrichKEGG",  #readable=TRUE,
                                          keyType = 'kegg',  #KEGG 富集
                                          organism='mmu',#"rno",
                                          pvalueCutoff = 0.05 #指定 p 值阈值（可指定 1 以输出全部
      )
      
      p=clusterProfiler::dotplot(object = xx,showCategory = 20,
                                 label_format =100) 
      p=p+ theme(axis.text.x = element_text(angle = 90, 
                                            vjust = 0.5, hjust=0.5))
      p
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-GO_enrichment--3.pdf'),plot = p,width = 13,height = 40,limitsize = F)
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-GO_enrichment--3.png'),plot = p,width = 13,height = 40,limitsize = F)
      
      xx
      write.csv(xx,file = paste0(avg_log2FC,"compareCluster-GO_enrichment.csv"))
      
      p=clusterProfiler::dotplot(gg,showCategory = 20,
                                 label_format = 40) 
      p4=p+ theme(axis.text.x = element_text(angle = 90, 
                                             vjust = 0.5, hjust=0.5))
      p4
      print(paste("保存位置",getwd(),sep = "  :   "))
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-KEGG_enrichment-2.pdf'),plot = p4,width = 13,height = 25,limitsize = F)
      ggsave(paste0(avg_log2FC,'_degs_compareCluster-KEGG_enrichment-2.png'),plot = p4,width = 13,height = 25,limitsize = F)
      
      gg
      openxlsx::write.xlsx(gg,file = paste0(avg_log2FC,"_compareCluster-KEGG_enrichment.xlsx"))
      
      getwd()
      openxlsx::write.xlsx(sce.markers,file = paste0(avg_log2FC,"_sce.markers_for_each_clusterfor_enrichment.xlsx"))
      
      
      #   save(xx.BP,xx.CC,xx.MF, gg, file = "~/silicosis/spatial_transcriptomicsharmony_cluster_0.5res_gsea/xx.Rdata")
      
      xx.BP@compareClusterResult$oncology="BP"
      xx.CC@compareClusterResult$oncology="CC"
      xx.MF@compareClusterResult$oncology="MF"
      
      
      xx.all=do.call(rbind,list(xx.BP@compareClusterResult,
                                xx.CC@compareClusterResult,
                                xx.MF@compareClusterResult))
      head(xx.all)    
      openxlsx::write.xlsx(xx.all,file = paste0(avg_log2FC,"_enrichments_all.xlsx"))    
      
      # save(xx.BP,xx.CC,xx.MF,xx.all,sce.markers,merged_degs, gg, file = "./xx.Rdata")
      
      result <- tryCatch({
        save(xx.BP,xx.CC,xx.MF,xx.all,sce.markers,merged_degs, gg, file = paste0(avg_log2FC,"_xx.Rdata"))
        # 在这里添加你希望在没有报错时执行的代码
        print("没有报错")
      }, error = function(e) {
        print(getwd())  # 在报错时执行的代码
        
        # 在这里添加你希望在报错时执行的额外代码
        save(xx.BP,xx.CC,xx.MF,xx.all,sce.markers, gg, file = paste0(avg_log2FC,"_xx.Rdata"))
      })
      
      # print(result)
      
      
      if (F) { 
        #大鼠
        print("===========开始go============")
        xx <-clusterProfiler::compareCluster(gcSample, fun="enrichGO",OrgDb="org.Rn.eg.db",
                                             readable=TRUE,
                                             ont = 'ALL',  #GO Ontology，可选 BP、MF、CC，也可以指定 ALL 同时计算 3 者
                                             pvalueCutoff=0.05) #organism="hsa", #'org.Hs.eg.db',
        
        print("===========开始 kegg============")
        gg<-clusterProfiler::compareCluster(gcSample,fun = "enrichKEGG",
                                            keyType = 'kegg',  #KEGG 富集
                                            organism="rno",
                                            pvalueCutoff = 0.05 #指定 p 值阈值（可指定 1 以输出全部
        )
        
        p=dotplot(xx) 
        p2=p+ theme(axis.text.x = element_text(angle = 90, 
                                               vjust = 0.5, hjust=0.5))
        p2
        ggsave('degs_compareCluster-GO_enrichment-2.pdf',plot = p2,width = 6,height = 20,limitsize = F)
        xx
        openxlsx::write.xlsx(xx,file = "compareCluster-GO_enrichment.xlsx")
        #  
        p=dotplot(gg) 
        p4=p+ theme(axis.text.x = element_text(angle = 90, 
                                               vjust = 0.5, hjust=0.5))
        p4
        print(paste("保存位置",getwd(),sep = "  :   "))
        ggsave('degs_compareCluster-KEGG_enrichment-2.pdf',plot = p4,width = 6,height = 12,limitsize = F)
        gg
        openxlsx::write.xlsx(gg,file = "compareCluster-KEGG_enrichment.xlsx")   
      }  
      
      setwd("../")
      
    }
    
    
  }
  
  
}

#######gsea----
{
  
  
  
  
}


###############3gsea-----------
#https://zhuanlan.zhihu.com/p/518144716
#https://yulab-smu.top/biomedical-knowledge-mining-book/faq.html

# library(org.Hs.eg.db)
library(org.Mm.eg.db)
library(clusterProfiler)
library(enrichplot)
library(tidyverse)
library(ggstatsplot)
{
  getwd()
  setwd("../")
  print(getwd())
  
  dir.create("5.GSEA_kegg_go")
  setwd("5.GSEA_kegg_go")
  
  ## 物种设置
  organism = 'mmu'    #  人类'hsa' 小鼠'mmu'   
  OrgDb = 'org.Mm.eg.db'#人类"org.Hs.eg.db" 小鼠"org.Mm.eg.db"
  
  head(specialized_degs)
  rownames(specialized_degs)=specialized_degs$gene
  
  #### 按照需要可选择不同的DEG方法数据集 ####
  need_DEG <- specialized_degs %>%
    dplyr::filter(p_val_adj<0.05)
  need_DEG <- need_DEG[,c(2,5)] #选择log2FoldChange和pvalue（凑成数据框）
  
  colnames(need_DEG) <- c('log2FoldChange','pvalue')
  need_DEG$SYMBOL <- rownames(need_DEG)
  
  ##### 创建gsea分析的geneList（包含从大到小排列的log2FoldChange和ENTREZID信息）####
  #转化id  
  df <- bitr(rownames(need_DEG), 
             fromType = "SYMBOL",
             toType =  "ENTREZID",
             OrgDb = OrgDb) #人数据库org.Hs.eg.db 小鼠org.Mm.eg.db
  need_DEG <- merge(need_DEG, df, by='SYMBOL')  #按照SYMBOL合并注释信息
  head(need_DEG) # https://yulab-smu.top/biomedical-knowledge-mining-book/faq.html
  geneList <- need_DEG$log2FoldChange
  names(geneList) <- need_DEG$ENTREZID
  geneList <- sort(geneList, decreasing = T)   #从大到小排序
head(geneList)
  

##### gsea富集 ####
KEGG_kk_entrez <- gseKEGG(geneList     = geneList,
                          organism     = organism, #人hsa 鼠mmu
                          pvalueCutoff = 0.25)  #实际为padj阈值,可调整 
KEGG_kk <- DOSE::setReadable(KEGG_kk_entrez, 
                             OrgDb=OrgDb,
                             keyType='ENTREZID')#转化id             

GO_kk_entrez <- gseGO(geneList     = geneList,
                      ont          = "ALL",  # "BP"、"MF"和"CC"或"ALL"
                      OrgDb        = OrgDb,#人类org.Hs.eg.db 鼠org.Mm.eg.db
                      keyType      = "ENTREZID",
                      pvalueCutoff = 0.25)   #实际为padj阈值可调整
GO_kk <- DOSE::setReadable(GO_kk_entrez, 
                           OrgDb=OrgDb,
                           keyType='ENTREZID')#转化id 
print(getwd())
save(KEGG_kk_entrez, GO_kk_entrez, file = "dega_GSEA_result.RData")


#首先对富集结果进行条件筛选，一般认为|NES|>1，NOM pvalue<0.05，FDR（padj）<0.25的通路是显著富集的；还可以从结果中细分出上下调通路单独绘图，以下代码仅展示KEGG通路富集结果的上调通路。 gseaplot2(）函数既可以对单独的通路绘图，也可以合并几个通路一起绘图；
KEGG_kk@result
kegg_gsea_res=data.frame(KEGG_kk@result)
##选取富集结果
kk_gse <- KEGG_kk
kk_gse_entrez <- KEGG_kk_entrez

###条件筛选 
#一般认为|NES|>1，NOM pvalue<0.05，FDR（padj）<0.25的通路是显著富集的
kk_gse_cut <- kk_gse[kk_gse$pvalue<0.05 & kk_gse$p.adjust<0.25 & abs(kk_gse$NES)>1]
kk_gse_cut_down <- kk_gse_cut[kk_gse_cut$NES < 0,]
kk_gse_cut_up <- kk_gse_cut[kk_gse_cut$NES > 0,]

#选择展现NES前几个通路 
down_gsea <- kk_gse_cut_down[tail(order(kk_gse_cut_down$NES,decreasing = T),100),]
up_gsea <- kk_gse_cut_up[head(order(kk_gse_cut_up$NES,decreasing = T),100),]
diff_gsea <- kk_gse_cut[head(order(abs(kk_gse_cut$NES),decreasing = T),100),]


#### 经典的GSEA图 uo----- 
up_gsea$Description

i=2
gselist=list()
for (i in 1:length(up_gsea$Description)) {
 
   gseap1 <- gseaplot2(kk_gse,
                      up_gsea$ID[i],#富集的ID编号
                      title = up_gsea$Description[i],#标题
                      color = "red", #GSEA线条颜色
                      base_size = 20,#基础字体大小
                      rel_heights = c(1.5, 0.5, 1),#副图的相对高度
                      subplots = 1:3,   #要显示哪些副图 如subplots=c(1,3) #只要第一和第三个图
                      ES_geom = "line", #enrichment score用线还是用点"dot"
                      pvalue_table = T) #显示pvalue等信息
   gselist[[up_gsea$Description[i]]]=gseap1 
  print(paste0(getwd(),up_gsea$Description[i]))
  
pdf(file =  paste0(up_gsea$Description[i], '_GSEA_up_1.pdf'), width =12, height =8)  
print(gseap1)   
  dev.off()
  
}


pdf(file =  paste0("all_split", '_GSEA_up_1.pdf'), width =12, height =8)  
for (i in names(gselist)) {
  print(gselist[[i]]) 
}
dev.off()

###下调gsea------
{
  
 down_gsea$Description
  
  i=2
  gselist=list()
  for (i in 1:length(down_gsea$Description)) {
    
    gseap1 <- gseaplot2(kk_gse,
                        down_gsea$ID[i],#富集的ID编号
                        title = down_gsea$Description[i],#标题
                        color = "red", #GSEA线条颜色
                        base_size = 20,#基础字体大小
                        rel_heights = c(1.5, 0.5, 1),#副图的相对高度
                        subplots = 1:3,   #要显示哪些副图 如subplots=c(1,3) #只要第一和第三个图
                        ES_geom = "line", #enrichment score用线还是用点"dot"
                        pvalue_table = T) #显示pvalue等信息
    gselist[[down_gsea$Description[i]]]=gseap1 
    print(paste0(getwd(),down_gsea$Description[i]))
    
    pdf(file =  paste0(down_gsea$Description[i], '_GSEA_down_1.pdf'), width =12, height =8)  
    print(gseap1)   
    dev.off()
    
  }
  
  
  
  pdf(file =  paste0("all_split", '_GSEA_down_1.pdf'), width =12, height =8)  
  for (i in names(gselist)) {
    print(gselist[[i]]) 
  }
  dev.off()
  
}

#### 合并 GSEA通路 
gseap2 <- gseaplot2(kk_gse,
                    up_gsea$ID,#富集的ID编号
                    title = "UP_GSEA_all",#标题
                    color = "red",#GSEA线条颜色
                    base_size = 20,#基础字体大小
                    rel_heights = c(1.5, 0.5, 1),#副图的相对高度
                    subplots = 1:3, #要显示哪些副图 如subplots=c(1,3) #只要第一和第三个图
                    ES_geom = "line",#enrichment score用线还是用点"dot"
                    pvalue_table = T) #显示pvalue等信息
ggsave(gseap2, filename = "combined_GSEA_up_all.pdf",width =12,height =12)


go_gsea_res=data.frame(GO_kk@result)

}