From 24b1cc244005177ab06464daee97351d277de3e2 Mon Sep 17 00:00:00 2001
From: Tanushree Tunstall <tanu@tunstall.in>
Date: Fri, 18 Sep 2020 11:55:08 +0100
Subject: [PATCH] saving work

---
 scripts/plotting/corr_adjusted_PS_LIG.R | 87 +++++++++++++++++++++----
 1 file changed, 76 insertions(+), 11 deletions(-)

diff --git a/scripts/plotting/corr_adjusted_PS_LIG.R b/scripts/plotting/corr_adjusted_PS_LIG.R
index 8f02a70..ed96b4b 100644
--- a/scripts/plotting/corr_adjusted_PS_LIG.R
+++ b/scripts/plotting/corr_adjusted_PS_LIG.R
@@ -14,6 +14,7 @@ getwd()
 source("Header_TT.R")
 require(cowplot)
 source("combining_dfs_plotting.R")
+source("my_pairs_panel.R")
 # should return the following dfs, directories and variables
 
 # PS combined: 
@@ -82,7 +83,58 @@ rm( merged_df2, merged_df2_comp, merged_df2_lig, merged_df2_comp_lig, my_df_u, m
 #===========================
 table(df_ps$duet_outcome)
 
+
+#===========================
+# Data for Correlation plots:foldx
+#===========================
+#============================
+# adding foldx scaled values
+# scale data b/w -1 and 1
+#============================
+n = which(colnames(df_ps) == "ddg"); n 
+
+my_min = min(df_ps[,n]); my_min 
+my_max = max(df_ps[,n]); my_max 
+
+df_ps$foldx_scaled = ifelse(df_ps[,n] < 0
+                            , df_ps[,n]/abs(my_min)
+                            , df_ps[,n]/my_max) 
+# sanity check
+my_min = min(df_ps$foldx_scaled); my_min 
+my_max = max(df_ps$foldx_scaled); my_max
+
+if (my_min == -1 && my_max == 1){
+  cat("PASS: foldx ddg successfully scaled b/w -1 and 1"
+      , "\nProceeding with assigning foldx outcome category")
+}else{
+  cat("FAIL: could not scale foldx ddg values"
+      , "Aborting!")
+}
+
+
+#================================
+# adding foldx outcome category
+# ddg<0 = "Stabilising" (-ve)
+#=================================
+
+c1 = table(df_ps$ddg < 0)
+df_ps$foldx_outcome = ifelse(df_ps$ddg < 0, "Stabilising", "Destabilising")
+c2 = table(df_ps$ddg < 0)
+
+if ( all(c1 == c2) ){
+  cat("PASS: foldx outcome successfully created")
+}else{
+  cat("FAIL: foldx outcome could not be created. Aborting!")
+  exit()
+}
+
+table(df_ps$foldx_outcome)
+
+
+#======================
 # adding log cols 
+#======================
+ 
 df_ps$log10_or_mychisq = log10(df_ps$or_mychisq)
 df_ps$neglog_pval_fisher = -log10(df_ps$pval_fisher)
 
@@ -92,14 +144,21 @@ df_ps$neglog_pwald_kin = -log10(df_ps$pwald_kin)
 # subset data to generate pairwise correlations
 cols_to_select =  c("duet_scaled"
                     
-                    , "log10_or_mychisq"
-                    , "neglog_pval_fisher"
+                    , "foldx_scaled"
                     
-                    #, "or_kin"
-                    #, "neglog_pwald_kin"
+                    #, "log10_or_mychisq"
+                    #, "neglog_pval_fisher"
+                    
+                    , "or_kin"
+                    , "neglog_pwald_kin"
                     
                     , "af"
                     
+                    , "asa"
+                    , "rsa"
+                    , "kd_values"
+                    , "rd_values"
+                    
                     , "duet_outcome"
                     , drug)
 
@@ -113,14 +172,20 @@ dim(corr_data_ps)
 # assign nice colnames (for display)
 my_corr_colnames = c("DUET"
                      
-                     , "Log(OR)"
-                     , "-Log(P)"
+                     , "Foldx"
+                     #, "Log(OR)"
+                     #, "-Log(P)"
                      
-                     #, "OR adjusted"
-                     #, "-Log(P wald)"
+                     , "OR adjusted"
+                     , "-Log(P wald)"
                      
                      , "AF"
                      
+                     , "ASA"
+                     , "RSA"
+                     , "KD"
+                     , "RD"
+                     
                      , "duet_outcome"
                      , drug)
 
@@ -161,9 +226,9 @@ OutPlot1 = pairs.panels(my_corr_ps[1:(length(my_corr_ps)-1)]
              , jitter = T
              #, alpha = .05
              #, points(pch = 19, col = c("#f8766d", "#00bfc4"))
-             , cex = 3
-             , cex.axis = 2.5
-             , cex.labels = 2.1
+             , cex = 2
+             , cex.axis = 1.5
+             , cex.labels = 1.5
              , cex.cor = 1
              , smooth = F
 )