From a6d93b3fa8542305ce01e693ee3cb48169d96ee4 Mon Sep 17 00:00:00 2001
From: Tanushree Tunstall <tanu@tunstall.in>
Date: Tue, 9 Aug 2022 21:55:24 +0100
Subject: [PATCH] starting corr plots

---
 scripts/functions/corr_plot_data.R            |  24 ---
 .../plotting_thesis/corr_plots_thesis.R       | 175 +++++++++---------
 2 files changed, 92 insertions(+), 107 deletions(-)

diff --git a/scripts/functions/corr_plot_data.R b/scripts/functions/corr_plot_data.R
index 970910f..e6aff38 100644
--- a/scripts/functions/corr_plot_data.R
+++ b/scripts/functions/corr_plot_data.R
@@ -1,28 +1,7 @@
 #!/usr/bin/env Rscript  
 #########################################################
 # TASK: Script to format data for Correlation plots: 
-
 # corr_data_extract()
-# INPUT:
-    # df: data with all parameters (my_use case)
-        # merged_df3 or merged_df2!?
-    # gene: [sanity check]
-    # drug: relates to a column name that will need to extracted
-    # ligand_dist_colname = LigDist_colname (variable from plotting_globals()
-   
-# colnames_to_extract = c("mutationinformation", "duet_affinity_change")
-# display_colnames_key = c(mutationinformation = "MUT" , duet_affinity_change = "DUET")
-# extract_scaled_cols = T or F, so that parameters with the _scaled suffix can be extracted.
-    # NOTE*: No formatting applied to these cols i.e display name
-
-# RETURNS: DF
-    # containing all the columns required for generating downstream correlation plots
-
-# TODO: ADD
-#lineage_count_all
-#lineage_count_unique
-#my_df['lineage_proportion']      = my_df['lineage_count_unique']/my_df['lineage_count_all']
-#my_df['dist_lineage_proportion'] = my_df['lineage_count_unique']/total_mtblineage_uc
 
 ##################################################################
 # LigDist_colname   #from globals: plotting_globals.R
@@ -31,14 +10,11 @@
 corr_data_extract <- function(df
                               , gene
                               , drug
-                              #, ligand_dist_colname = LigDist_colname
                               , colnames_to_extract
                               , colnames_display_key
                               , extract_scaled_cols = F){
   
   if ( missing(colnames_to_extract) || missing(colnames_display_key) ){
-  #if ( missing(colnames_to_extract) ){
-      
     cat("\n=========================================="
         , "\nCORR PLOTS data: ALL params"
         , "\n=========================================")
diff --git a/scripts/plotting/plotting_thesis/corr_plots_thesis.R b/scripts/plotting/plotting_thesis/corr_plots_thesis.R
index 1470076..348b8cb 100644
--- a/scripts/plotting/plotting_thesis/corr_plots_thesis.R
+++ b/scripts/plotting/plotting_thesis/corr_plots_thesis.R
@@ -127,48 +127,50 @@ dev.off()
 #================
 # ppi2 affinity
 #================
-corr_ppi2_colnames = c("mCSM-PPI2"
-                       , "MAF"             
-                       , "Log(OR)"       
-                       , "-Log(P)"
-                       , "PPI-Dist"  # "interface_dist"
-                       , "dst_mode"
-                       , drug)
-
-corr_ppi2_colnames%in%colnames(corr_plotdf)
-corr_df_ppi2 = corr_plotdf[, corr_ppi2_colnames]
-corr_df_ppi2 = corr_df_ppi2[corr_df_ppi2["PPI-Dist"]<DistCutOff,]
-complete_obs_ppi2 = nrow(corr_df_ppi2) - sum(is.na(corr_df_ppi2$`Log(OR)`))
-cat("\nComplete muts for ppi2 affinity for", gene, ":", complete_obs_ppi2)
-
-color_coln = which(colnames(corr_df_ppi2) == "dst_mode")
-end = which(colnames(corr_df_ppi2) == drug)
-ncol_omit = 3 #omit dist col
-corr_end = end-ncol_omit
-
-#------------------------
-# Output: ppi2 corrP
-#------------------------
-corr_ppi2P =  paste0(outdir_images
-                    ,tolower(gene)
-                    ,"_corr_ppi2.svg" )
-
-cat("Corr plot ppi2 with coloured dots:", corr_ppi2P)
-svg(corr_ppi2P, width = 10, height = 10)
-
-my_corr_pairs(corr_data_all = corr_df_ppi2
-              , corr_cols = colnames(corr_df_ppi2[1:corr_end])
-              , corr_method = "spearman"
-              , colour_categ_col = colnames(corr_df_ppi2[color_coln]) #"dst_mode"
-              , categ_colour =  c("red", "blue")
-              , density_show = F
-              , hist_col = "coral4"
-              , dot_size = 2
-              , ats = 1.5
-              , corr_lab_size = 3
-              , corr_value_size = 1)
-
-dev.off()
+if (tolower(gene)%in%geneL_ppi2){
+  corr_ppi2_colnames = c("mCSM-PPI2"
+                         , "MAF"             
+                         , "Log(OR)"       
+                         , "-Log(P)"
+                         , "PPI-Dist"  # "interface_dist"
+                         , "dst_mode"
+                         , drug)
+  
+  corr_ppi2_colnames%in%colnames(corr_plotdf)
+  corr_df_ppi2 = corr_plotdf[, corr_ppi2_colnames]
+  corr_df_ppi2 = corr_df_ppi2[corr_df_ppi2["PPI-Dist"]<DistCutOff,]
+  complete_obs_ppi2 = nrow(corr_df_ppi2) - sum(is.na(corr_df_ppi2$`Log(OR)`))
+  cat("\nComplete muts for ppi2 affinity for", gene, ":", complete_obs_ppi2)
+  
+  color_coln = which(colnames(corr_df_ppi2) == "dst_mode")
+  end = which(colnames(corr_df_ppi2) == drug)
+  ncol_omit = 3 #omit dist col
+  corr_end = end-ncol_omit
+  
+  #------------------------
+  # Output: ppi2 corrP
+  #------------------------
+  corr_ppi2P =  paste0(outdir_images
+                      ,tolower(gene)
+                      ,"_corr_ppi2.svg" )
+  
+  cat("Corr plot ppi2 with coloured dots:", corr_ppi2P)
+  svg(corr_ppi2P, width = 10, height = 10)
+  
+  my_corr_pairs(corr_data_all = corr_df_ppi2
+                , corr_cols = colnames(corr_df_ppi2[1:corr_end])
+                , corr_method = "spearman"
+                , colour_categ_col = colnames(corr_df_ppi2[color_coln]) #"dst_mode"
+                , categ_colour =  c("red", "blue")
+                , density_show = F
+                , hist_col = "coral4"
+                , dot_size = 2
+                , ats = 1.5
+                , corr_lab_size = 3
+                , corr_value_size = 1)
+  
+  dev.off()
+}
 
 # FIXME: ADD distance
 #==================
@@ -177,48 +179,52 @@ dev.off()
 #================
 # NA affinity
 #================
-corr_na_colnames = c("mCSM-NA"
-                     , "MAF"             
-                     , "Log(OR)"       
-                     , "-Log(P)"
-                     , "NA-Dist"  # "NA_dist"
-                     , "dst_mode"
-                     , drug)
+if (tolower(gene)%in%geneL_na){
+    
+  corr_na_colnames = c("mCSM-NA"
+                       , "MAF"             
+                       , "Log(OR)"       
+                       , "-Log(P)"
+                       , "NA-Dist"  # "NA_dist"
+                       , "dst_mode"
+                       , drug)
+  
+  corr_na_colnames%in%colnames(corr_plotdf)
+  corr_df_na = corr_plotdf[, corr_na_colnames]
+  corr_df_na = corr_df_na[corr_df_na["NA-Dist"]<DistCutOff,]
+  complete_obs_na = nrow(corr_df_na) - sum(is.na(corr_df_na$`Log(OR)`))
+  cat("\nComplete muts for NA affinity for", gene, ":", complete_obs_na)
+  
+  color_coln = which(colnames(corr_df_na) == "dst_mode")
+  end = which(colnames(corr_df_na) == drug)
+  ncol_omit = 3 #omit dist col
+  corr_end = end-ncol_omit
+  
+  #------------------------
+  # Output: mCSM-NA corrP
+  #------------------------
+  corr_naP =  paste0(outdir_images
+                     ,tolower(gene)
+                     ,"_corr_na.svg" )
+  
+  cat("Corr plot mCSM-NA with coloured dots:", corr_naP)
+  svg(corr_naP, width = 10, height = 10)
+  
+  my_corr_pairs(corr_data_all = corr_df_na
+                , corr_cols = colnames(corr_df_na[1:corr_end])
+                , corr_method = "spearman" 
+                , colour_categ_col = colnames(corr_df_na[color_coln]) #"dst_mode"
+                , categ_colour =  c("red", "blue")
+                , density_show = F
+                , hist_col = "coral4"
+                , dot_size = 2
+                , ats = 1.5
+                , corr_lab_size = 3
+                , corr_value_size = 1)
+  
+  dev.off()
+}
 
-corr_na_colnames%in%colnames(corr_plotdf)
-corr_df_na = corr_plotdf[, corr_na_colnames]
-corr_df_na = corr_df_na[corr_df_na["NA-Dist"]<DistCutOff,]
-complete_obs_na = nrow(corr_df_na) - sum(is.na(corr_df_na$`Log(OR)`))
-cat("\nComplete muts for NA affinity for", gene, ":", complete_obs_na)
-
-color_coln = which(colnames(corr_df_na) == "dst_mode")
-end = which(colnames(corr_df_na) == drug)
-ncol_omit = 3 #omit dist col
-corr_end = end-ncol_omit
-
-#------------------------
-# Output: mCSM-NA corrP
-#------------------------
-corr_naP =  paste0(outdir_images
-                   ,tolower(gene)
-                   ,"_corr_na.svg" )
-
-cat("Corr plot mCSM-NA with coloured dots:", corr_naP)
-svg(corr_naP, width = 10, height = 10)
-
-my_corr_pairs(corr_data_all = corr_df_na
-              , corr_cols = colnames(corr_df_na[1:corr_end])
-              , corr_method = "spearman" 
-              , colour_categ_col = colnames(corr_df_na[color_coln]) #"dst_mode"
-              , categ_colour =  c("red", "blue")
-              , density_show = F
-              , hist_col = "coral4"
-              , dot_size = 2
-              , ats = 1.5
-              , corr_lab_size = 3
-              , corr_value_size = 1)
-
-dev.off()
 
 ####################################################
 # CONSERVATION
@@ -265,3 +271,6 @@ my_corr_pairs(corr_data_all = corr_df_cons
               , corr_value_size = 1)
 
 dev.off()
+
+
+# 
\ No newline at end of file