From 2909f49eecc0c7dc780230c5c63ac85b5ed2c936 Mon Sep 17 00:00:00 2001
From: Costa <47926492N@ICO.SCS.local>
Date: Mon, 28 Mar 2022 12:49:29 +0200
Subject: [PATCH] =?UTF-8?q?A=C3=B1adido=20plot=20para=20expresi=C3=B3n=20d?=
 =?UTF-8?q?e=20genes.?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 BDAccess/app.R | 65 +++++++++++++++++++++++++++++++++++++++++---------
 1 file changed, 54 insertions(+), 11 deletions(-)

diff --git a/BDAccess/app.R b/BDAccess/app.R
index b4b4c0c..31a7971 100644
--- a/BDAccess/app.R
+++ b/BDAccess/app.R
@@ -68,12 +68,16 @@ ui <- fluidPage(
                  sidebarPanel(
                      textInput("sqlquery", label = "sqlquery", value = ""),
                      uiOutput("PATID"),
-                     checkboxInput("cd45_chk", "Purificación CD45")
+                     checkboxInput("sct_sel", "Mostrar filtrados"),
+                     checkboxInput("cd45_chk", "Purificación CD45"),
+                     textInput("genes", label="genes", value = "")
                  ),
                  mainPanel(
                      tabsetPanel(
                          tabPanel("Table", tableOutput("sc_table")),
-                         tabPanel("Plots", plotOutput("sc_plot"))
+                         tabPanel("Plots", 
+                                  plotOutput("sc_plot", height = "1000px"),
+                                  plotOutput("sc_expr"), height = "600px")
                      )
                  )
         )
@@ -701,20 +705,25 @@ server <- function(input, output) {
     })
     
     output$sc_table<-renderTable({
-        if (input$sqlquery != ""){
-            print(input$sqlquery)
-            sqlQuery(dta, input$sqlquery)
-        }else{
-            if (!is.null(input$sc_cod)){
-                sqlFetch(dta, "CNAG") %>% filter(CODIGO %in% input$sc_cod)
+        if (input$sct_sel){
+            if (input$sqlquery != ""){
+                print(input$sqlquery)
+                sqlQuery(dta, input$sqlquery)
             }else{
-                sqlFetch(dta, "CNAG")
+                if (!is.null(input$sc_cod)){
+                    sqlFetch(dta, "CNAG") %>% filter(CODIGO %in% input$sc_cod)
+                }else{
+                    sqlFetch(dta, "CNAG")
+                }
             }
+        }else{
+            sqlFetch(dta, "CNAG")
         }
+
     })
     
     output$sc_plot <-renderPlot({
-        meta<-readRDS(paste0(scRNAseqRoute,"metadata_full_object.rds"))
+        meta<<-readRDS(paste0(scRNAseqRoute,"metadata_full_object.rds"))
         
         if (input$sqlquery != ""){
             sc_codigos<-sqlQuery(dta, input$sqlquery) %>% pull(CNAG_NAME)
@@ -742,8 +751,42 @@ server <- function(input, output) {
             spread(predicted.id, N) %>% gather("predicted.id","N",-sample) %>% mutate(N=case_when(is.na(N)~0,TRUE~N))
         g2<-ggheatmap(meta_perc, "sample","predicted.id", "N", color = "grey50")
         ggpubr::ggarrange(g1,g2, ncol=1, heights = c(0.3, 0.7))
-    }, height = 1000)
+    })
+
+    output$sc_expr <-renderPlot({
+        if (input$genes != ""){
+            expr<-readRDS(paste0(scRNAseqRoute,"expression_full_object.rds"))
+            genes<-strsplit(input$genes, ",")[[1]]
 
+            if (length(genes) > 1){
+                df.expr<-as.data.frame(as.matrix(expr[genes,]))
+                df.expr["Gene"]<-rownames(df.expr)
+                mdf.expr<-melt(df.expr)
+            }else{
+                df.expr<-as.data.frame(t(as.matrix(expr[genes[1],])))
+                df.expr["Gene"]<-genes[1]
+                mdf.expr<-melt(df.expr)
+            }
+            
+            alldata<-merge(meta, mdf.expr, by.x="barcode", by.y = "variable")
+            
+        }
+        
+        # order <- clustsort(alldata %>% spread(Gene, value) %>% select(predicted.id, all_of(genes)) %>%  
+        #                        group_by(predicted.id) %>% summarise(across(all_of(genes), mean)) %>%  as.data.frame)
+        # 
+        # g1<-ggplot(alldata, aes(predicted.id, value, fill=predicted.id))+
+        #     geom_violin(scale = "width")+
+        #     geom_jitter(width=0.2, size=0.1, alpha=0.3)+
+        #     scale_x_discrete(limits=order$x)+
+        #     guides(fill=F)+
+        #     facet_wrap(.~Gene)+
+        #     theme_bw()+
+        #     theme(axis.text.x = element_text(angle=90, hjust=1, vjust = 0.5))
+        g2<-ggheatmap(alldata, x="predicted.id",y="Gene",value="value", color="grey")+coord_equal()
+        # ggpubr::ggarrange(g1,g2, ncol=1)
+        g2
+    })
 }
 
 # Run the application