From e0f14ed266a6891de17af672767eba8da23fd75c Mon Sep 17 00:00:00 2001
From: Tanushree Tunstall <tanu@tunstall.in>
Date: Wed, 26 Aug 2020 16:39:10 +0100
Subject: [PATCH] sorted subcols_axis script to generate correct axis cols for
 both PS and lig plots

---
 scripts/plotting/Header_TT.R                 |   0
 scripts/plotting/barplot_colour_function.R   |   0
 scripts/plotting/barplots_subcolours_PS.R    |   0
 scripts/plotting/barplots_subcolours_aa_PS.R |  47 +++---
 scripts/plotting/basic_barplots_PS.R         |   2 +-
 scripts/plotting/combining_two_df_FIXME.R    |   0
 scripts/plotting/mcsm_mean_stability.R       |   0
 scripts/plotting/plotting_data.R             |   2 +-
 scripts/plotting/subcols_axis_PS.R           | 147 ++++++++++++-------
 9 files changed, 117 insertions(+), 81 deletions(-)
 mode change 100644 => 100755 scripts/plotting/Header_TT.R
 mode change 100644 => 100755 scripts/plotting/barplot_colour_function.R
 mode change 100644 => 100755 scripts/plotting/barplots_subcolours_PS.R
 mode change 100644 => 100755 scripts/plotting/barplots_subcolours_aa_PS.R
 mode change 100644 => 100755 scripts/plotting/basic_barplots_PS.R
 mode change 100644 => 100755 scripts/plotting/combining_two_df_FIXME.R
 mode change 100644 => 100755 scripts/plotting/mcsm_mean_stability.R
 mode change 100644 => 100755 scripts/plotting/plotting_data.R
 mode change 100644 => 100755 scripts/plotting/subcols_axis_PS.R

diff --git a/scripts/plotting/Header_TT.R b/scripts/plotting/Header_TT.R
old mode 100644
new mode 100755
diff --git a/scripts/plotting/barplot_colour_function.R b/scripts/plotting/barplot_colour_function.R
old mode 100644
new mode 100755
diff --git a/scripts/plotting/barplots_subcolours_PS.R b/scripts/plotting/barplots_subcolours_PS.R
old mode 100644
new mode 100755
diff --git a/scripts/plotting/barplots_subcolours_aa_PS.R b/scripts/plotting/barplots_subcolours_aa_PS.R
old mode 100644
new mode 100755
index 84b5811..5850abf
--- a/scripts/plotting/barplots_subcolours_aa_PS.R
+++ b/scripts/plotting/barplots_subcolours_aa_PS.R
@@ -1,3 +1,4 @@
+#!/usr/bin/env Rscript   
 getwd()
 setwd("~/git/LSHTM_analysis/scripts/plotting")
 getwd()
@@ -42,14 +43,30 @@ cat(paste0("Variables imported:"
            , "\nAngstrom symbol:", angstroms_symbol))
 
 # clear excess variable
-rm(my_df, upos, dup_muts, my_df_u_lig)
+rm(dup_muts_cols, mut_pos_cols_lig, my_df_cols, my_df_u_cols_lig, upos)
+
 #=======================================================================
+# !!! very important!!!!
 #================
 # Inspecting mut_pos_cols
-# position numbers and colours
+# position numbers and colours and assigning axis colours based on lab_fg
+# of the correct df
 # open file from desktop ("sample_axis_cols") for cross checking
 #================
+# very important!
+#my_axis_colours = mut_pos_cols$lab_fg
 
+if ( nrow(mut_pos_cols) == length(unique(my_df_u_cols$position)) ){
+  print("PASS: lengths checked, assigning axis colours")
+  my_axis_colours = mut_pos_cols$lab_fg
+  cat("length of axis colours:", length(my_axis_colours)
+      , "\nwhich corresponds to the number of positions on the x-axis of the plot")
+}else{
+  print("FAIL:lengths mismatch, could not assign axis colours")
+  quit()
+}
+
+# further sanity checks
 table(mut_pos_cols$lab_bg)
 check_lab_bg = sum( table(mut_pos_cols$lab_bg) ) == nrow(mut_pos_cols) # should  be True
 check_lab_bg    
@@ -70,12 +87,6 @@ if (check_lab_bg && check_lab_bg2 && check_lab_fg) {
   quit()
 }
 
-# very important!
-my_axis_colours = mut_pos_cols$lab_fg
-
-# now clear mut_pos_cols
-rm(mut_pos_cols)
-
 #=======
 # output
 #=======
@@ -89,13 +100,13 @@ plot_bp_aa_subcols_duet = paste0(plotdir, "/", bp_aa_subcols_duet)
 # Data for plots
 #================
 # REASSIGNMENT as necessary
-df  = my_df_u
+df  = my_df_u_cols
 
 # sanity checks
 str(df)
 
 ###########################
-# 2: Plot: DUET scores
+# Plot: DUET scores
 ###########################
 
 #==========================
@@ -137,7 +148,7 @@ snp_count = sort(unique(snpsBYpos_df$snpsBYpos))
 if (is.factor(df$duet_outcome)){
   print("duet_outcome is factor")
 }else{
-  print("convert duet_outcome to factor")
+  print("converting duet_outcome to factor")
   df$duet_outcome = as.factor(df$duet_outcome)
 }
 
@@ -165,25 +176,17 @@ tapply(df$duet_scaled, df$duet_outcome, max)
 u = unique(df$duet_scaled) 
 cat("No. of unique values in normalised data:", length(u))
 
-#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
-# Run this section if rounding is to be used
-# specify number for rounding
-#n = 3 
-#df$duet_scaledR = round(df$duet_scaled, n)  
-#ur = unique(df$duet_scaledR)
-
-# create an extra column called group which contains the "gp name and score" 
+# Define group
+# Create an extra column called group which contains the "gp name and score" 
 # so colours can be generated for each unique values in this column
-
-#my_grp = df$duet_scaledR # rounding
 my_grp = df$duet_scaled # no rounding
-#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 df$group <- paste0(df$duet_outcome, "_", my_grp, sep = "") 
 
 # Call the function to create the palette based on the group defined above
 colours <- ColourPalleteMulti(df, "duet_outcome", "my_grp")
 print(paste0("Colour palette generated for: ", length(colours), " colours"))
 my_title = "Protein stability (DUET)"
+cat("No. of axis colours: ", length(my_axis_colours))
 
 #========================
 # plot with axis colours
diff --git a/scripts/plotting/basic_barplots_PS.R b/scripts/plotting/basic_barplots_PS.R
old mode 100644
new mode 100755
index c28332b..47d34d9
--- a/scripts/plotting/basic_barplots_PS.R
+++ b/scripts/plotting/basic_barplots_PS.R
@@ -32,7 +32,7 @@ cat(paste0("Directories imported:"
            , "\ndrug:", drug
            , "\ngene:", gene
            , "\ngene_match:", gene_match
-           , "\nLength of upos:", length(upos))
+           , "\nLength of upos:", length(upos)
            , "\nAngstrom symbol:", angstroms_symbol))       
            
 # clear excess variable
diff --git a/scripts/plotting/combining_two_df_FIXME.R b/scripts/plotting/combining_two_df_FIXME.R
old mode 100644
new mode 100755
diff --git a/scripts/plotting/mcsm_mean_stability.R b/scripts/plotting/mcsm_mean_stability.R
old mode 100644
new mode 100755
diff --git a/scripts/plotting/plotting_data.R b/scripts/plotting/plotting_data.R
old mode 100644
new mode 100755
index b9f2197..291579e
--- a/scripts/plotting/plotting_data.R
+++ b/scripts/plotting/plotting_data.R
@@ -67,7 +67,7 @@ cat("\nInput dimensions:", dim(my_df))
 #str(my_df)
 
 ###########################
-# extract unique mutations
+# extract unique mutation entries
 ###########################
 
 # check for duplicate mutations
diff --git a/scripts/plotting/subcols_axis_PS.R b/scripts/plotting/subcols_axis_PS.R
old mode 100644
new mode 100755
index 5fe0ab3..867d8bd
--- a/scripts/plotting/subcols_axis_PS.R
+++ b/scripts/plotting/subcols_axis_PS.R
@@ -1,7 +1,6 @@
 #########################################################
-# TASK: Adding colours to positions labels according to 
-# active site residues. This is so these can be seen promptly
-# when visualising the barplot.
+# TASK: Adding colours to dfs so they can be used for plotting
+# add  cols to each of the my_df* dfs
 #########################################################
 #=======================================================================
 getwd()
@@ -15,46 +14,45 @@ source("plotting_data.R")
    # my_df_u_lig
    # dup_muts
 
+cat(paste0("Directories imported:"
+           , "\ndatadir:", datadir
+           , "\nindir:", indir
+           , "\noutdir:", outdir
+           , "\nplotdir:", plotdir))
+
+cat(paste0("Variables imported:"
+           , "\ndrug:", drug
+           , "\ngene:", gene
+           , "\ngene_match:", gene_match
+           , "\nLength of upos:", length(upos)
+    , "\nAngstrom symbol:", angstroms_symbol))       
+
+# clear excess variable
+rm(upos, dup_muts, my_df_u, my_df_u_lig)
+
+# This is because we want to assign the colours to my_df
+# and then resubset accordingly for our plots to avoid multiple merges
+
 #=======================================================================
-###########################
-# Read file: struct params
-###########################
-cat("Reading struct params including mcsm:", in_filename_params)
+# df to use: my_df
+# NOTE: my_df contains duplicate muts but its ok as you are only adding 
+# colours to positions
 
-my_df = read.csv(infile_params
-                 #, stringsAsFactors = F
-                 , header = T)
+# sanity checks: ensure my_df is ordered by position: it should be 
+my_df$position; my_df$mutationinformation
 
-cat("Input dimensions:", dim(my_df)) 
+my_df_o = my_df[order(my_df$position),]
+my_df_o$position; my_df_o$mutationinformation
 
-# clear variables
-rm(in_filename_params, infile_params)
+head(my_df_o$position) == head(my_df$position)
+head(my_df_o$mutationinformation) == head(my_df$mutationinformation)
+tail(my_df_o$position) == tail(my_df$position)
+tail(my_df_o$mutationinformation) == tail(my_df$mutationinformation)
 
-# quick checks
-colnames(my_df)
-str(my_df)
+my_df = my_df_o
 
-# check for duplicate mutations
-if ( length(unique(my_df$mutationinformation)) != length(my_df$mutationinformation)){
-  cat(paste0("CAUTION:", " Duplicate mutations identified"
-             , "\nExtracting these..."))
-  dup_muts = my_df[duplicated(my_df$mutationinformation),]
-  dup_muts_nu = length(unique(dup_muts$mutationinformation))
-  cat(paste0("\nDim of duplicate mutation df:", nrow(dup_muts)
-             , "\nNo. of unique duplicate mutations:", dup_muts_nu
-             , "\n\nExtracting df with unique mutations only"))
-  my_df_u = my_df[!duplicated(my_df$mutationinformation),]
-}else{
-  cat(paste0("No duplicate mutations detected"))
-  my_df_u = my_df
-}
-
-upos = unique(my_df_u$position)
-cat("Dim of clean df:"); cat(dim(my_df_u))
-cat("\nNo. of unique mutational positions:"); cat(length(upos))
-#=======================================================================
 # create a new df with unique position numbers and cols
-position = unique(my_df$position) #130
+position = unique(my_df$position) 
 position_cols = as.data.frame(position)
 
 head(position_cols) ; tail(position_cols)
@@ -143,6 +141,7 @@ mut_pos_cols = merge(position_cols, aa_cols_ref
             , all.x = TRUE) 
 
 head(mut_pos_cols)
+
 # replace NA"s 
 # :column "lab_bg" with "white"
 # : column "lab_fg" with "black"
@@ -165,39 +164,69 @@ head(df0$position); tail(df0$position)
 head(df1$position); tail(df1$position)
 
 # should now have 3 extra columns
-my_df = merge(df0, df1
+my_df_cols = merge(df0, df1
               , by = "position"
               , all.x = TRUE)
 
 # sanity check
-my_df[my_df$position == "49",]
-my_df[my_df$position == "13",]
+my_df_cols[my_df_cols$position == "49",]
+my_df_cols[my_df_cols$position == "13",]
 
-rm(df0, df1)
-#===========
-# Merge 3: Merge mut_pos_cols with mcsm df_u
-# Now combined the positions with aa colours with 
-# the mcsm_data
-#===========
-# dfs to merge
-df0 = my_df_u # my_df_u
-df1 = mut_pos_cols
+###########################
+# extract unique mutation entries
+###########################
 
-# check the column on which merge will be performed
-head(df0$position); tail(df0$position)
-head(df1$position); tail(df1$position)
+# check for duplicate mutations
+if ( length(unique(my_df_cols$mutationinformation)) != length(my_df_cols$mutationinformation)){
+   cat(paste0("\nCAUTION:", " Duplicate mutations identified"
+              , "\nExtracting these..."))
+   dup_muts_cols = my_df_cols[duplicated(my_df_cols$mutationinformation),]
+   dup_muts_cols_nu = length(unique(dup_muts_cols$mutationinformation))
+   cat(paste0("\nDim of duplicate mutation df:", nrow(dup_muts_cols)
+              , "\nNo. of unique duplicate mutations:", dup_muts_cols_nu
+              , "\n\nExtracting df with unique mutations only"))
+   my_df_u_cols = my_df_cols[!duplicated(my_df_cols$mutationinformation),]
+}else{
+   cat(paste0("\nNo duplicate mutations detected"))
+   my_df_u_cols = my_df_cols
+}
+
+upos = unique(my_df_u_cols$position)
+cat("\nDim of clean df:"); cat(dim(my_df_u_cols))
+cat("\nNo. of unique mutational positions:"); cat(length(upos), "\n")
 
-# should now have 3 extra columns
-my_df_u = merge(df0, df1
-              , by = "position"
-              , all.x = TRUE)
 
 # sanity check
-my_df[my_df$position == "49",]
-my_df[my_df$position == "13",]
+my_df_u_cols[my_df_u_cols$position == "49",]
+my_df_u_cols[my_df_u_cols$position == "13",]
+my_df_u_cols[my_df_u_cols$position == "103",]
 
+###########################
+# extract mutations <10Angstroms
+###########################
+table(my_df_u_cols$ligand_distance<10)
+
+my_df_u_cols_lig = my_df_u_cols[my_df_u_cols$ligand_distance <10,]
+angstroms_symbol = "\u212b"
+cat(paste0("There are ", nrow(my_df_u_cols_lig), " sites lying within 10", angstroms_symbol, " of the ligand"))
+
+#=================
+# very important!
+#=================
+#my_axis_colours = mut_pos_cols$lab_fg # doesn't work if positions numbers are subsetted as in ligand
+# need the equivalent of the mut_pos_cols for ligand
+
+# get position numbers for ligand
+lig_pos = my_df_u_cols_lig$position
+
+# subset mut_pos_cols for ligand positions
+mut_pos_cols_lig = mut_pos_cols[mut_pos_cols$position %in% lig_pos,]
+#my_axis_colours = mut_pos_cols_lig$lab_fg
+
+#====================================================================
 # clear variables
 rm(aa_cols_ref
+   , my_df
    , df0
    , df1
    , position_cols
@@ -206,3 +235,7 @@ rm(aa_cols_ref
    , lab_fg
    , position)
 
+#######################################################################
+# end of script
+#######################################################################
+