HIF1_tagets_signature

0.1 Parameters

suffix = ""

0.2 functions

0.3 Data

xeno = readRDS("./Data/10x_xeno_1000.Rds")
lung = readRDS("./Data/lung_cancercells_withTP_onlyPatients.rds")
lung_patients = lung$patient.ident %>% unique() %>% as.character()
lung_patients_filtered = lung_patients[!(lung_patients %in% c("X1055new","X1099"))] # remove patients with less than 100 malignant cells
lung = subset(x = lung,subset = patient.ident %in% lung_patients_filtered)

1 HIF1 targets

#from:
# https://pubmed.ncbi.nlm.nih.gov/36384128/

hif_targets = "LDHA
PPFIA4
PFKFB4
AK4
ERO1A
BNIP3L
BNIP3
PFKL
ENO1
MIR210HG
TPI1
HILPDA
PGK1
INSIG2
AK4P1
AC114803.1
ENO2
FUT11
EGLN3
NDRG1
BNIP3P1
WDR54
OVOL1-AS1
STC2
DDX41
C4orf47
GYS1
ANKRD37
BICDL2
P4HA1
PDK1
OSMR
PKM
LDHAP5
AL158201.1
PFKP
MIR210
NLRP3P1
GSX1
AL109946.1
PGAM1
SLC16A3
ISM2
TCAF2
ARID3A
KDM3A
JMJD6
C4orf3
" 


hif_targets = strsplit(hif_targets,"\n")[[1]]
hif_targets

 [1] "LDHA"       "PPFIA4"     "PFKFB4"     "AK4"        "ERO1A"      "BNIP3L"     "BNIP3"      "PFKL"       "ENO1"      
[10] "MIR210HG"   "TPI1"       "HILPDA"     "PGK1"       "INSIG2"     "AK4P1"      "AC114803.1" "ENO2"       "FUT11"     
[19] "EGLN3"      "NDRG1"      "BNIP3P1"    "WDR54"      "OVOL1-AS1"  "STC2"       "DDX41"      "C4orf47"    "GYS1"      
[28] "ANKRD37"    "BICDL2"     "P4HA1"      "PDK1"       "OSMR"       "PKM"        "LDHAP5"     "AL158201.1" "PFKP"      
[37] "MIR210"     "NLRP3P1"    "GSX1"       "AL109946.1" "PGAM1"      "SLC16A3"    "ISM2"       "TCAF2"      "ARID3A"    
[46] "KDM3A"      "JMJD6"      "C4orf3"    

2 Xeno

hif_targets_exp = FetchData(object = xeno,vars = c(hif_targets))

Warning in FetchData.Seurat(object = xeno, vars = c(hif_targets)) : The following requested variables were not found: AK4P1, BNIP3P1, LDHAP5, AL158201.1, MIR210, NLRP3P1, AL109946.1

hif_targets_exp = hif_targets_exp[,colSums(hif_targets_exp[])>0] #remove no expression genes
hif_cor = cor(hif_targets_exp)
pht1 = pheatmap(mat = hif_cor,silent = T)

num_of_clusters = 3
clustering_distance = "euclidean"
myannotation = as.data.frame(cutree(pht1[["tree_row"]], k = num_of_clusters)) #split into k clusters
 
names(myannotation)[1] = "cluster"
  myannotation$cluster = as.factor(myannotation$cluster)
  
  palette1 <-brewer.pal(num_of_clusters, "Paired")

  names(palette1) = unique(myannotation$cluster)
  ann_colors = list (cluster = palette1)
  annotation = list(ann_colors = ann_colors, myannotation = myannotation)
  
  colors <- c(seq(-1,1,by=0.01))
  my_palette <- c("blue",colorRampPalette(colors = c("blue", "white", "red"))
                                                   (n = length(colors)-3), "red")


  print_tab(plt = 
                pheatmap(mat = hif_cor,annotation_col =  annotation[["myannotation"]], annotation_colors = annotation[["ann_colors"]], clustering_distance_rows = clustering_distance,clustering_distance_cols = clustering_distance,color = my_palette,breaks = colors,show_rownames = T,show_colnames = F,fontsize_row = 9)
            ,title = "genes expression heatmap")

genes expression heatmap

NA

cluster_3_genes = annotation[["myannotation"]] %>% dplyr::filter(cluster == 3) %>% rownames() #take relevant genes
hif_targets_by_tp = FetchData(object = xeno,vars = c(cluster_3_genes)) %>% rowSums() %>% as.data.frame() #mean expression
names(hif_targets_by_tp)[1] = "hif1_targets_mean"

hif_targets_by_tp = cbind(hif_targets_by_tp,FetchData(object = xeno,vars = c("orig.ident","treatment"))) # add vars

hif_targets_by_tp_forPlot =  hif_targets_by_tp %>% mutate(orig.ident = paste("model",orig.ident)) #add "model" before model num
my_comparisons = list( c("NT", "OSI") )
plt = ggplot(hif_targets_by_tp_forPlot, aes(x=treatment, y=hif1_targets_mean)) +
  geom_boxplot() +
  scale_y_continuous(limits = c(0, 37))+ #scale axis 
  facet_wrap(~orig.ident, nrow = 2, strip.position = "top")+
  stat_compare_means(comparisons = my_comparisons,method = "wilcox.test") + # Add pairwise comparisons p-value
ylab("HIF target genes")
xeno_cluster_3 = cluster_3_genes
print_tab(plt = plt,title = "cluster 3 genes")

cluster 3 genes

NA

cluster_3_genes = annotation[["myannotation"]] %>% dplyr::filter(cluster == 3 | cluster == 2) %>% rownames() #take relevant genes
hif_targets_by_tp = FetchData(object = xeno,vars = c(cluster_3_genes)) %>% rowSums() %>% as.data.frame() #mean expression
names(hif_targets_by_tp)[1] = "hif1_targets_mean"

hif_targets_by_tp = cbind(hif_targets_by_tp,FetchData(object = xeno,vars = c("orig.ident","treatment"))) # add vars

hif_targets_by_tp_forPlot =  hif_targets_by_tp %>% mutate(orig.ident = paste("model",orig.ident)) #add "model" before model num
my_comparisons = list( c("NT", "OSI") )
plt = ggplot(hif_targets_by_tp_forPlot, aes(x=treatment, y=hif1_targets_mean)) +
  geom_boxplot() +
  scale_y_continuous(limits = c(0, 50))+ #scale axis 
  facet_wrap(~orig.ident, nrow = 2, strip.position = "top")+
  stat_compare_means(comparisons = my_comparisons,method = "wilcox.test")+ # Add pairwise comparisons p-value
ylab("HIF target genes")
xeno_cluster_3_2 = cluster_3_genes

print_tab(plt = plt,title = "cluster 3+2 genes")

cluster 3+2 genes

NA

3 Patients

hif_targets_exp = FetchData(object = lung,vars = c(hif_targets))

Warning in FetchData.Seurat(object = lung, vars = c(hif_targets)) :
  The following requested variables were not found: AC114803.1, BNIP3P1, OVOL1-AS1, BICDL2, LDHAP5, AL158201.1, NLRP3P1, GSX1, AL109946.1

hif_targets_exp = hif_targets_exp[,colSums(hif_targets_exp[])>0] #remove no expression genes
hif_cor = cor(hif_targets_exp)
pht1 = pheatmap(mat = hif_cor,silent = T)

num_of_clusters = 4
clustering_distance = "euclidean"
myannotation = as.data.frame(cutree(pht1[["tree_row"]], k = num_of_clusters)) #split into k clusters
 
names(myannotation)[1] = "cluster"
  myannotation$cluster = as.factor(myannotation$cluster)
  
  palette1 <-brewer.pal(num_of_clusters, "Paired")

  names(palette1) = unique(myannotation$cluster)
  ann_colors = list (cluster = palette1)
  annotation = list(ann_colors = ann_colors, myannotation = myannotation)
  
  colors <- c(seq(-1,1,by=0.01))
  my_palette <- c("blue",colorRampPalette(colors = c("blue", "white", "red"))
                                                   (n = length(colors)-3), "red")


  print_tab(plt = 
                pheatmap(mat = hif_cor,annotation_col =  annotation[["myannotation"]], annotation_colors = annotation[["ann_colors"]], clustering_distance_rows = clustering_distance,clustering_distance_cols = clustering_distance,color = my_palette,breaks = colors,show_rownames = T,show_colnames = F,fontsize_row = 9)
            ,title = "genes expression heatmap")

genes expression heatmap

NA

cluster_3_genes = annotation[["myannotation"]] %>% dplyr::filter(cluster == 1) %>% rownames() #take relevant genes
hif_targets_by_tp = FetchData(object = lung,vars = c(cluster_3_genes)) %>% rowSums() %>% as.data.frame() #mean expression
names(hif_targets_by_tp)[1] = "hif1_targets_mean"

hif_targets_by_tp = cbind(hif_targets_by_tp,FetchData(object = lung,vars = c("patient.ident","time.point"))) # add vars

hif_targets_by_tp_forPlot =  hif_targets_by_tp %>% mutate(patient.ident = paste("patient ",patient.ident)) #add "model" before model num
my_comparisons = list( c("pre-treatment", "on-treatment") )
plt = ggplot(hif_targets_by_tp_forPlot, aes(x=time.point, y=hif1_targets_mean)) +
  geom_boxplot() +
  scale_y_continuous(limits = c(0, max(hif_targets_by_tp_forPlot$hif1_targets_mean)*1.1))+ #scale axis
  facet_wrap(~patient.ident, nrow = 2, strip.position = "top")+
  stat_compare_means(comparisons = my_comparisons,method = "wilcox.test") + # Add pairwise comparisons p-value
ylab("HIF target genes")

print_tab(plt = plt,title = "cluster 1 genes")

cluster 1 genes

NA

cluster_3_genes = annotation[["myannotation"]] %>% dplyr::filter(cluster == 1 | cluster == 4) %>% rownames() #take relevant genes
hif_targets_by_tp = FetchData(object = lung,vars = c(cluster_3_genes)) %>% rowSums() %>% as.data.frame() #mean expression
names(hif_targets_by_tp)[1] = "hif1_targets_mean"

hif_targets_by_tp = cbind(hif_targets_by_tp,FetchData(object = lung,vars = c("patient.ident","time.point"))) # add vars

hif_targets_by_tp_forPlot =  hif_targets_by_tp %>% mutate(patient.ident = paste("patient ",patient.ident)) #add "model" before model num
my_comparisons = list( c("pre-treatment", "on-treatment") )
plt = ggplot(hif_targets_by_tp_forPlot, aes(x=time.point, y=hif1_targets_mean)) +
  geom_boxplot() +
  scale_y_continuous(limits = c(0, max(hif_targets_by_tp_forPlot$hif1_targets_mean)*1.1))+ #extend y axis for pval
  facet_wrap(~patient.ident, nrow = 2, strip.position = "top")+
  stat_compare_means(comparisons = my_comparisons,method = "wilcox.test") + # Add pairwise comparisons p-value
ylab("HIF target genes")

print_tab(plt = plt,title = "cluster 1+4 genes")

cluster 1+4 genes

NA

cluster_3_genes =xeno_cluster_3
hif_targets_by_tp = FetchData(object = lung,vars = c(cluster_3_genes)) %>% rowSums() %>% as.data.frame() #mean expression
names(hif_targets_by_tp)[1] = "hif1_targets_mean"

hif_targets_by_tp = cbind(hif_targets_by_tp,FetchData(object = lung,vars = c("patient.ident","time.point"))) # add vars

hif_targets_by_tp_forPlot =  hif_targets_by_tp %>% mutate(patient.ident = paste("patient ",patient.ident)) #add "model" before model num
my_comparisons = list( c("pre-treatment", "on-treatment") )
plt = ggplot(hif_targets_by_tp_forPlot, aes(x=time.point, y=hif1_targets_mean)) +
  geom_boxplot() +
  scale_y_continuous(limits = c(0, max(hif_targets_by_tp_forPlot$hif1_targets_mean)*1.1))+ #extend y axis for pval
  facet_wrap(~patient.ident, nrow = 2, strip.position = "top")+
  stat_compare_means(comparisons = my_comparisons,method = "wilcox.test") + # Add pairwise comparisons p-value
ylab("HIF target genes")

print_tab(plt = plt,title = "models cluster 3")

models cluster 3

NA

cluster_3_genes =xeno_cluster_3_2
hif_targets_by_tp = FetchData(object = lung,vars = c(cluster_3_genes)) %>% rowSums() %>% as.data.frame() #mean expression

Warning in FetchData.Seurat(object = lung, vars = c(cluster_3_genes)) : The following requested variables were not found: AC114803.1, OVOL1-AS1, BICDL2

names(hif_targets_by_tp)[1] = "hif1_targets_mean"

hif_targets_by_tp = cbind(hif_targets_by_tp,FetchData(object = lung,vars = c("patient.ident","time.point"))) # add vars

hif_targets_by_tp_forPlot =  hif_targets_by_tp %>% mutate(patient.ident = paste("patient ",patient.ident)) #add "model" before model num
my_comparisons = list( c("pre-treatment", "on-treatment") )
plt = ggplot(hif_targets_by_tp_forPlot, aes(x=time.point, y=hif1_targets_mean)) +
  geom_boxplot() +
  scale_y_continuous(limits = c(0, max(hif_targets_by_tp_forPlot$hif1_targets_mean)*1.1))+ #extend y axis for pval
  facet_wrap(~patient.ident, nrow = 2, strip.position = "top")+
  stat_compare_means(comparisons = my_comparisons,method = "wilcox.test") + # Add pairwise comparisons p-value
ylab("HIF target genes")

print_tab(plt = plt,title = "models cluster 3+2")

models cluster 3+2

NA