#Pathway radar plot
library(data.table)library(fgsea)library(ggplot2)library(msigdbr)library(dplyr)library(tidyverse)library(fmsb)
m_df <- msigdbr(species = "Homo sapiens", category = "C2",subcategory = "CP:KEGG")fgsea_sets<- m_df %>% split(x = .$gene_symbol, f = .$gs_name)
markers <- read.csv("~/tumoursubset__0.2_RNA_rPCA.csv", header = TRUE, check.names = FALSE) %>% as.data.frame()

markers$cluster[markers$cluster=="0"]<-"Resting Memory B cells"markers$cluster[markers$cluster=="1"]<-"Activated Memory B cells"markers$cluster[markers$cluster=="2"]<-"Conventional Plasma cells"markers$cluster[markers$cluster=="3"]<-"Naive B cells"markers$cluster[markers$cluster=="4"]<-"Stressed Plasma cells"markers$cluster[markers$cluster=="5"]<-"Atypical Memory B cells"markers$cluster[markers$cluster=="6"]<-"IGKC-high Plasma/Plasmablasts"markers$cluster[markers$cluster=="7"]<-"Transitional/Proliferative 1"markers$cluster[markers$cluster=="8"]<-"Transitional/Proliferative 2"markers$cluster[markers$cluster=="9"]<-"MT1X-high Plasma/Plasmablasts"
colnames(markers) [1] <- "gene1" 
fgseaRes<-data.frame()for(i in c("Naive B cells", "Activated Memory B cells", "Resting Memory B cells", "IGKC-high Plasma/Plasmablasts", "Conventional Plasma cells", "Stressed Plasma cells", "Atypical Memory B cells", "Transitional/Proliferative 1", "Transitional/Proliferative 2", "MT1X-high Plasma/Plasmablasts")){  cluster.genes<- markers %>% dplyr::filter(cluster == i) %>%arrange(desc(avg_log2FC)) %>% dplyr::select(gene,avg_log2FC)   ranks<- deframe(cluster.genes)  fgseaRes1<- fgsea(fgsea_sets, stats = ranks, nperm = 1000)   fgseaRes1$cluster<-i  fgseaRes<-bind_rows(fgseaRes, fgseaRes1)}


top5<- fgseaRes %>%  filter(pval < 0.05)%>%  group_by(cluster) %>%  slice_max(n = 5, order_by = NES)


pathway<-as.data.frame(unique(top5$pathway))
colnames(pathway)<-"pathway"
for(i in c("Naive B cells", "Activated Memory B cells", "Resting Memory B cells", "IGKC-high Plasma/Plasmablasts", "Conventional Plasma cells", "Stressed Plasma cells", "Atypical Memory B cells", "Transitional/Proliferative 1", "Transitional/Proliferative 2", "MT1X-high Plasma/Plasmablasts")){  cluster.pathway<- fgseaRes %>% dplyr::filter(cluster == i)%>%dplyr::select(pathway,NES)  pathway<-left_join(pathway, cluster.pathway,by ="pathway")}

colnames(pathway)[-1]<-c("Naive B cells", "Activated Memory B cells", "Resting Memory B cells", "IGKC-high Plasma/Plasmablasts", "Conventional Plasma cells", "Stressed Plasma cells", "Atypical Memory B cells", "Transitional/Proliferative 1", "Transitional/Proliferative 2", "MT1X-high Plasma/Plasmablasts")

rownames(pathway)<-pathway$pathway
pathway[is.na(pathway)] <- 0
pathway <- data.frame(t(pathway[-1]))
#Add rows with highest and lowest valuespathway <- rbind(rep(2.2374599,25) , rep(-2.9921065,25) , pathway)
#Rename pathwayspathway <- pathway %>%   rename(    TCR_Signalling = KEGG_T_CELL_RECEPTOR_SIGNALING_PATHWAY,    NK_mediated_cytotoxicity = KEGG_NATURAL_KILLER_CELL_MEDIATED_CYTOTOXICITY,    VEGF_Signalling= KEGG_VEGF_SIGNALING_PATHWAY,    IgA_production= KEGG_INTESTINAL_IMMUNE_NETWORK_FOR_IGA_PRODUCTION,    GVHD = KEGG_GRAFT_VERSUS_HOST_DISEASE,    Cell_adhesion= KEGG_CELL_ADHESION_MOLECULES_CAMS,    T1D = KEGG_TYPE_I_DIABETES_MELLITUS,    Asthma = KEGG_ASTHMA,    Protein_export = KEGG_PROTEIN_EXPORT,    Vibrio_Cholerae_Infection = KEGG_VIBRIO_CHOLERAE_INFECTION,    Antigen_presentation = KEGG_ANTIGEN_PROCESSING_AND_PRESENTATION,    N_Glycan_Biosynthesis = KEGG_N_GLYCAN_BIOSYNTHESIS,    Cardiac_muscle_contraction = KEGG_CARDIAC_MUSCLE_CONTRACTION,    Insulin_signalling =KEGG_INSULIN_SIGNALING_PATHWAY,    MTOR_signalling =KEGG_MTOR_SIGNALING_PATHWAY,    Ribosome = KEGG_RIBOSOME,    MAPK_signalling = KEGG_MAPK_SIGNALING_PATHWAY,    Endocytosis = KEGG_ENDOCYTOSIS,    Prion_Diseases = KEGG_PRION_DISEASES,    Spliceosome = KEGG_SPLICEOSOME,    Lupus =KEGG_SYSTEMIC_LUPUS_ERYTHEMATOSUS,    Oxidative_Phosphorylation = KEGG_OXIDATIVE_PHOSPHORYLATION,    Parkinsons = KEGG_PARKINSONS_DISEASE,    Viral_Mycarditis = KEGG_VIRAL_MYOCARDITIS,    Leishmania = KEGG_LEISHMANIA_INFECTION,    Allograft_rejection =KEGG_ALLOGRAFT_REJECTION,    Autoimmune_thyroid = KEGG_AUTOIMMUNE_THYROID_DISEASE  )
# Select relevant columnspathway <- pathway[,c(1,2,3,4,7,10,12,13,17,18,19,22)] 
#Plot Radar Plotradarchart(pathway, plwd = 5, pcol = colors_border) +   legend(x=1.5, y=1, legend = rownames(pathway[-c(1,2),]), bty = "n", pch=20, text.col = "black", col=colors_border, cex=1, pt.cex=3)
colors_border= c("#9E9E9E","#D64550",                 "#FF977E","#A43B76",                 "#9A64A0","#750985",                 "#5ECBC8", "#28788D", "#37A794",                  "#4C5D8A", "#039BE5", "#BF360C")