Pathfinder (enrichment ComplexHeatmap)

dev-notebooks/STRINGdb-Analysis.Rmd

@@ -3,8 +3,8 @@ title: 'STRINGdb: A Database of Known and Predicted Protein-protein Interactions
 author: "Co-adapted by Haowen He and Rachael White from the 2015 Vignette by Andrea Franceschini"
 subtitle: IDEA Alzheimer's Data Analysis Bootcamp - Summer 2022
 output:
-  html_document: default
   pdf_document: default
+  html_document: default
 always_allow_html: true
 ---
 
@@ -180,12 +180,12 @@ string_db$plot_network(hits, payload_id=payload_id)
 ## Clustering
 
 ```{r, fig.width=10, fig.height=10, fig.fullwidth=TRUE, fig.fullheight=TRUE}
-clustersList <- string_db$get_clusters(example1_mapped$STRING_id[1:600])
+#clustersList <- string_db$get_clusters(example1_mapped$STRING_id[1:600])
 # plot first 4 clusters
-par(mfrow=c(2,2))
-for(i in seq(1:4)){
-   string_db$plot_network(clustersList[[i]])
-   }
+#par(mfrow=c(2,2))
+#for(i in seq(1:4)){
+#   string_db$plot_network(clustersList[[i]])
+#   }
 ```
 
 \newpage
@@ -237,14 +237,10 @@ dim(seurat_mapped)
 hits_seurat <- seurat_mapped$STRING_id[1:200]
 ```
 
-\newpage
-
 ```{r, fig.width=10, fig.height=10, fig.fullwidth=TRUE, fig.fullheight=TRUE}
-string_db$plot_network(hits_seurat)
+#string_db$plot_network(hits_seurat)
 ```
 
-\newpage
-
 ```{r}
 # PAYLOAD MECHANISM
 # filter by p-value and add a color column
@@ -272,22 +268,175 @@ payload_id_seurat <- string_db$post_payload( seurat_mapped_pval05$STRING_id,
                                       colors = seurat_mapped_pval05$color )
 ```
 
-\newpage
-
 ```{r, fig.width=10, fig.height=10, fig.fullwidth=TRUE, fig.fullheight=TRUE}
-string_db$plot_network(hits_seurat)
+#string_db$plot_network(hits_seurat)
 ```
 
 ```{r, , fig.width=10, fig.height=10, fig.fullwidth=TRUE, fig.fullheight=TRUE}
 # display a STRING network png with the "halo"
 string_db$plot_network(hits_seurat, payload_id = payload_id_seurat)
+title(sub = "STRINGdb: protein-protein interaction of Astrocytes at 2mo", 
+      cex.sub = 1,   font.sub = 3, col.sub = "darkgreen"
+      )
 
-clustersList_seurat <- string_db$get_clusters(seurat_mapped$STRING_id[1:600])
+# clustersList_seurat <- string_db$get_clusters(seurat_mapped$STRING_id[1:600])
 # plot first 4 clusters
-par(mfrow=c(2,2))
-for(i in seq(1:4)){
-   string_db$plot_network(clustersList_seurat[[i]])
-   }
+#par(mfrow=c(2,2))
+#for(i in seq(1:4)){
+#   string_db$plot_network(clustersList_seurat[[i]])
+#}
 ```
 
+\newpage
+
+```{r, , fig.width=10, fig.height=10, fig.fullwidth=TRUE, fig.fullheight=TRUE}
+# specify your cell type 
+my_celltype <- "Ast"
+# specify time point of interest
+my_timept_2 <- "4mo"
+
+# Load Seurat data
+# Change the filepath in quotes "" below to match where you 
+# have your celltype data saved
+# The expected filepath is "~/AlzheimersDS/student-notebooks/FinalMergedData-my_celltype.rds"
+# (where `my-celltype` is your celltype code)
+my_celltype_data <- readRDS("/data/Alzheimers_DS/FinalMergedData-Ast.rds")
+
+# Subset my celltype-specific Seurat object by time
+my_seurat_2 <- subset(x = my_celltype_data, subset = Age == my_timept_2)
+
+# set active identities of my subset seurat to "Mt" attribute (denoting variant)
+# for DE testing
+Idents(my_seurat_2) <- "Mt"
+summary(as.factor(my_seurat_2$Mt))
+
+# Run D.E. test to identify genes differentially expressed between
+# between diseased and normal cells, for my celltype and at this timepoint
+marker_genes.df_2 <- FindMarkers(my_seurat_2, 
+                               ident.1 = "V337M", ident.2 = "V337V") 
+
+# order results by p-value
+marker_genes.df_2 <- marker_genes.df_2 %>% arrange(p_val) 
+head(marker_genes.df_2)
+
+# format a new dataframe using the contents of
+# marker_genes.df, to pass to STRINGdb
+stringdb.df_2 <- marker_genes.df_2 %>% 
+  tibble::rownames_to_column("geneIDs") %>% 
+  dplyr::select(geneIDs, avg_log2FC, p_val)
+colnames(stringdb.df_2) <- c("geneIDs","logFC","pvalue")
+
+# mapping
+seurat_mapped_2 <- string_db$map(stringdb.df_2, "geneIDs", removeUnmappedRows = TRUE )
+dim(seurat_mapped_2)
+
+#View(seurat_mapped)
+hits_seurat_2 <- seurat_mapped_2$STRING_id[1:200]
+
+# PAYLOAD MECHANISM
+# filter by p-value and add a color column
+# (i.e. green down-regulated gened and red for up-regulated genes)
+head(subset(seurat_mapped_2, pvalue < 0.05), 20)
+
+dim(seurat_mapped_2[seurat_mapped_2$pvalue < 0.05,])
+
+seurat_mapped_pval05_2 <- string_db$add_diff_exp_color(subset(seurat_mapped_2, pvalue<0.05),
+                                                        logFcColStr="logFC" )
+head(seurat_mapped_pval05_2)
+
+table(seurat_mapped_pval05_2$color)
+
+dim(seurat_mapped_pval05_2)
+
+seurat_mapped_pval05_2[seurat_mapped_pval05_2$color == '#D7FFD7FF', ]
+
+seurat_mapped_pval05_2[seurat_mapped_pval05_2$color == '#FFEDEDFF', ]
+
+seurat_mapped_pval05_2[seurat_mapped_pval05_2$color == '#FFF6F6FF', ]
 
+# post payload information to the STRING server
+payload_id_seurat_2 <- string_db$post_payload( seurat_mapped_pval05_2$STRING_id,
+                                      colors = seurat_mapped_pval05_2$color )
+
+string_db$plot_network(hits_seurat_2, payload_id = payload_id_seurat_2)
+title(sub = "STRINGdb: protein-protein interaction of Astrocytes at 4mo", 
+      cex.sub = 1,   font.sub = 3, col.sub = "darkgreen"
+      )
+```
+
+\newpage
+
+```{r, , fig.width=10, fig.height=8, fig.fullwidth=TRUE, fig.fullheight=TRUE}
+# specify your cell type 
+my_celltype <- "Ast"
+# specify time point of interest
+my_timept_3 <- "6mo"
+
+# Load Seurat data
+# Change the filepath in quotes "" below to match where you 
+# have your celltype data saved
+# The expected filepath is "~/AlzheimersDS/student-notebooks/FinalMergedData-my_celltype.rds"
+# (where `my-celltype` is your celltype code)
+my_celltype_data <- readRDS("/data/Alzheimers_DS/FinalMergedData-Ast.rds")
+
+# Subset my celltype-specific Seurat object by time
+my_seurat_3 <- subset(x = my_celltype_data, subset = Age == my_timept_3)
+
+# set active identities of my subset seurat to "Mt" attribute (denoting variant)
+# for DE testing
+Idents(my_seurat_3) <- "Mt"
+summary(as.factor(my_seurat_3$Mt))
+
+# Run D.E. test to identify genes differentially expressed between
+# between diseased and normal cells, for my celltype and at this timepoint
+marker_genes.df_3 <- FindMarkers(my_seurat_3, 
+                               ident.1 = "V337M", ident.2 = "V337V") 
+
+# order results by p-value
+marker_genes.df_3 <- marker_genes.df_3 %>% arrange(p_val) 
+head(marker_genes.df_3)
+
+# format a new dataframe using the contents of
+# marker_genes.df, to pass to STRINGdb
+stringdb.df_3 <- marker_genes.df_3 %>% 
+  tibble::rownames_to_column("geneIDs") %>% 
+  dplyr::select(geneIDs, avg_log2FC, p_val)
+colnames(stringdb.df_3) <- c("geneIDs","logFC","pvalue")
+
+# mapping
+seurat_mapped_3 <- string_db$map(stringdb.df_3, "geneIDs", removeUnmappedRows = TRUE )
+dim(seurat_mapped_3)
+
+#View(seurat_mapped)
+hits_seurat_3 <- seurat_mapped_3$STRING_id[1:200]
+
+# PAYLOAD MECHANISM
+# filter by p-value and add a color column
+# (i.e. green down-regulated gened and red for up-regulated genes)
+head(subset(seurat_mapped_3, pvalue < 0.05), 20)
+
+dim(seurat_mapped_3[seurat_mapped_3$pvalue < 0.05,])
+
+seurat_mapped_pval05_3 <- string_db$add_diff_exp_color(subset(seurat_mapped_3, pvalue<0.05),
+                                                        logFcColStr="logFC" )
+head(seurat_mapped_pval05_3)
+
+table(seurat_mapped_pval05_3$color)
+
+dim(seurat_mapped_pval05_3)
+
+seurat_mapped_pval05_3[seurat_mapped_pval05_3$color == '#D7FFD7FF', ]
+
+seurat_mapped_pval05_3[seurat_mapped_pval05_3$color == '#FFEDEDFF', ]
+
+seurat_mapped_pval05_3[seurat_mapped_pval05_3$color == '#FFF6F6FF', ]
+
+# post payload information to the STRING server
+payload_id_seurat_3 <- string_db$post_payload( seurat_mapped_pval05_3$STRING_id,
+                                      colors = seurat_mapped_pval05_3$color )
+
+string_db$plot_network(hits_seurat_3, payload_id = payload_id_seurat_3)
+title(sub = "STRINGdb: protein-protein interaction of Astrocytes at 6mo", 
+      cex.sub = 1,   font.sub = 3, col.sub = "darkgreen"
+      )
+```