From dc5b5e2f11d8bf207dc8386faff577f03977b469 Mon Sep 17 00:00:00 2001 From: Tanushree Tunstall Date: Fri, 10 Sep 2021 18:20:45 +0100 Subject: [PATCH] added shorter scripts for each different processing for plots to make it wasire to read code --- scripts/plotting/coloured_bp_data.R | 80 +++ scripts/plotting/corr_data.R | 67 +++ scripts/plotting/dm_om_data.R | 416 ++++++++++++++++ scripts/plotting/logo_data.R | 142 ++++++ scripts/plotting/redundant/other_dfs_data.R | 117 +++++ scripts/plotting/redundant/other_plots_data.R | 470 ++++++++++++++++++ 6 files changed, 1292 insertions(+) create mode 100644 scripts/plotting/coloured_bp_data.R create mode 100644 scripts/plotting/corr_data.R create mode 100644 scripts/plotting/dm_om_data.R create mode 100644 scripts/plotting/logo_data.R create mode 100644 scripts/plotting/redundant/other_dfs_data.R create mode 100755 scripts/plotting/redundant/other_plots_data.R diff --git a/scripts/plotting/coloured_bp_data.R b/scripts/plotting/coloured_bp_data.R new file mode 100644 index 0000000..a1f0964 --- /dev/null +++ b/scripts/plotting/coloured_bp_data.R @@ -0,0 +1,80 @@ +#!/usr/bin/env Rscript +################################################################# +# TASK: Script to add bp colours ~ barplot heatmap +################################################################# + +my_df = merged_df3 + +cols_to_select = c("mutationinformation", "drtype" + , "wild_type" + , "position" + , "mutant_type" + , "chain", "ligand_id", "ligand_distance" + , "duet_stability_change", "duet_outcome", "duet_scaled" + , "ligand_affinity_change", "ligand_outcome", "affinity_scaled" + , "ddg_foldx", "foldx_scaled", "foldx_outcome" + , "deepddg", "deepddg_outcome" # comment out as not available for pnca + , "asa", "rsa", "rd_values", "kd_values" + , "af", "or_mychisq", "pval_fisher" + , "or_fisher", "or_logistic", "pval_logistic" + , "wt_prop_water", "mut_prop_water", "wt_prop_polarity", "mut_prop_polarity" + , "wt_calcprop", "mut_calcprop") + +#======================= +# Data for sub colours +# barplot: PS +#======================= + +cat("\nNo. of cols to select:", length(cols_to_select)) + +subcols_df_ps = my_df[, cols_to_select] + +cat("\nNo of unique positions for ps:" + , length(unique(subcols_df_ps$position))) + +# add count_pos col that counts the no. of nsSNPS at a position +setDT(subcols_df_ps)[, pos_count := .N, by = .(position)] + +# should be a factor +if (is.factor(subcols_df_ps$duet_outcome)){ + cat("\nDuet_outcome is factor") + table(subcols_df_ps$duet_outcome) +}else{ + cat("\nConverting duet_outcome to factor") + subcols_df_ps$duet_outcome = as.factor(subcols_df_ps$duet_outcome) + table(subcols_df_ps$duet_outcome) +} + +# should be -1 and 1 +min(subcols_df_ps$duet_scaled) +max(subcols_df_ps$duet_scaled) + +tapply(subcols_df_ps$duet_scaled, subcols_df_ps$duet_outcome, min) +tapply(subcols_df_ps$duet_scaled, subcols_df_ps$duet_outcome, max) + +# check unique values in normalised data +cat("\nNo. of unique values in duet scaled, no rounding:" + , length(unique(subcols_df_ps$duet_scaled))) + +# No rounding +my_grp = subcols_df_ps$duet_scaled; length(my_grp) + +# Add rounding is to be used +n = 3 +subcols_df_ps$duet_scaledR = round(subcols_df_ps$duet_scaled, n) + +cat("\nNo. of unique values in duet scaled", n, "places rounding:" + , length(unique(subcols_df_ps$duet_scaledR))) + +my_grp_r = subcols_df_ps$duet_scaledR # rounding + +# Add grp cols +subcols_df_ps$group <- paste0(subcols_df_ps$duet_outcome, "_", my_grp, sep = "") +subcols_df_ps$groupR <- paste0(subcols_df_ps$duet_outcome, "_", my_grp_r, sep = "") + +# Call the function to create the palette based on the group defined above +subcols_ps <- ColourPalleteMulti(subcols_df_ps, "duet_outcome", "my_grp") +subcolsR_ps <- ColourPalleteMulti(subcols_df_ps, "duet_outcome", "my_grp_r") + +cat("Colour palette generated for my_grp: ", length(subcols_ps), " colours") +cat("Colour palette generated for my_grp_r: ", length(subcolsR_ps), " colours") diff --git a/scripts/plotting/corr_data.R b/scripts/plotting/corr_data.R new file mode 100644 index 0000000..d33efc5 --- /dev/null +++ b/scripts/plotting/corr_data.R @@ -0,0 +1,67 @@ +#!/usr/bin/env Rscript +######################################################### +# TASK: Script to format data for corr plots +######################################################### + +#================================================= +# Data for Corrplots +#================================================= +cat("\n==========================================" + , "\nCORR PLOTS data: ALL params" + , "\n=========================================") + +# use data +#merged_df2 + +#---------------------------- +# columns for corr plots:PS +#---------------------------- +# NOTE: you can add mcsm_ppi column as well, and it will only select what it can find! +big_df_colnames = data.frame(names(merged_df2)) + +corr_cols_select <- c("mutationinformation", drug, "mutation_info_labels" + , "duet_stability_change", "ligand_affinity_change", "ddg_foldx", "asa", "rsa" + , "rd_values", "kd_values", "log10_or_mychisq", "neglog_pval_fisher","af" + , "deepddg", "ddg_dynamut", "ddg_dynamut2", "mcsm_na_affinity" + , "ddg_encom", "dds_encom", "ddg_mcsm", "ddg_sdm", "ddg_duet", "ligand_distance") + +#=========================== +# Corr data for plots: PS +# big_df ps: ~ merged_df2 +#=========================== + +corr_df_m2 = merged_df2[,colnames(merged_df2)%in%corr_cols_select] + +#=========================== +# Corr data for plots: PS +# short_df ps: ~merged_df3 +#=========================== + +corr_df_m3 = corr_df_m2[!duplicated(corr_df_m2$mutationinformation),] + +na_or = sum(is.na(corr_df_m3$log10_or_mychisq)) +check1 = nrow(corr_df_m3) - na_or; check1 + +if (nrow(corr_df_m3) == nrow(merged_df3) && nrow(merged_df3_comp) == check1) { + cat( "\nPASS: No. of rows for corr_df_m3 match" + , "\nPASS: No. of OR values checked: " , check1) +} else { + cat("\nFAIL: Numbers mismatch:" + , "\nExpected nrows: ", nrow(merged_df3) + , "\nGot: ", nrow(corr_df_m3) + , "\nExpected OR values: ", nrow(merged_df3_comp) + , "\nGot: ", check1) +} + +cat("\nCorr Data created:" +, "\n===================================" +, "\ncorr_df_m2: created from merged_df2" +, "\n===================================" +, "\nnrows:", nrow(corr_df_m2) +, "\nncols:", ncol(corr_df_m2) +, "\n===================================" +, "\ncorr_df_m3: created from merged_df3" +, "\n===================================" +, "\nnrows:", nrow(corr_df_m3) +, "\nncols:", ncol(corr_df_m3) +) diff --git a/scripts/plotting/dm_om_data.R b/scripts/plotting/dm_om_data.R new file mode 100644 index 0000000..4bd82e7 --- /dev/null +++ b/scripts/plotting/dm_om_data.R @@ -0,0 +1,416 @@ +#!/usr/bin/env Rscript +######################################################### +# TASK: Script to format data for dm om plots: +# generating LF data +# sourced by get_plotting_dfs.R +######################################################### +##======================================================================== +# cols to select: +# THINK: whu + +comb_df <- merged_df3[, c("mutationinformation", "mutation" + , "mutation_info","mutation_info_labels" + , "position" + , LigDist_colname + , "duet_stability_change", "duet_scaled", "duet_outcome" + , "ligand_affinity_change", "affinity_scaled", "ligand_outcome" + , "ddg_foldx", "foldx_scaled", "foldx_outcome" + , "deepddg", "deepddg_scaled", "deepddg_outcome" + , "asa", "rsa" + , "rd_values", "kd_values" + , "log10_or_mychisq", "neglog_pval_fisher", "af" + , "mcsm_na_affinity", "mcsm_na_scaled", "mcsm_na_outcome" + , "ddg_dynamut", "ddg_dynamut_scaled","ddg_dynamut_outcome" + , "ddg_encom", "ddg_encom_scaled", "ddg_encom_outcome" + , "dds_encom", "dds_encom_scaled", "dds_encom_outcome" + , "ddg_mcsm", "ddg_mcsm_scaled", "ddg_mcsm_outcome" + , "ddg_sdm", "ddg_sdm_scaled", "ddg_sdm_outcome" + , "ddg_duet", "ddg_duet_scaled", "ddg_duet_outcome" + , "ddg_dynamut2","ddg_dynamut2_scaled", "ddg_dynamut2_outcome")] + + +comb_df_s = arrange(comb_df, position) + +#======================================================================= +fact_cols = colnames(comb_df_s)[grepl( "_outcome|_info", colnames(comb_df_s) )] +fact_cols +lapply(comb_df_s[, fact_cols], class) +comb_df_s[, fact_cols] <- lapply(comb_df_s[, fact_cols], as.factor) + +if (any(lapply(comb_df_s[, fact_cols], class) == "character")){ + cat("\nChanging cols to factor") + comb_df_s[, fact_cols] <- lapply(comb_df_s[, fact_cols],as.factor) + if (all(lapply(comb_df_s[, fact_cols], class) == "factor")){ + cat("\nSuccessful: cols changed to factor") + } +} +lapply(comb_df_s[, fact_cols], class) + +#======================================================================= +table(comb_df_s$mutation_info) + + # further checks to make sure dr and other muts are indeed unique +dr_muts = comb_df_s[comb_df_s$mutation_info == dr_muts_col,] +dr_muts_names = unique(dr_muts$mutation) + +other_muts = comb_df_s[comb_df_s$mutation_info == other_muts_col,] +other_muts_names = unique(other_muts$mutation) + +if ( table(dr_muts_names%in%other_muts_names)[[1]] == length(dr_muts_names) && + table(other_muts_names%in%dr_muts_names)[[1]] == length(other_muts_names) ){ + cat("PASS: dr and other muts are indeed unique") +}else{ + cat("FAIL: dr and others muts are NOT unique!") + quit() +} + +# pretty display names i.e. labels to reduce major code duplication later +foo_cnames = data.frame(colnames(comb_df_s)) +names(foo_cnames) <- "old_name" + +stability_suffix <- paste0(delta_symbol, delta_symbol, "G") +flexibility_suffix <- paste0(delta_symbol, delta_symbol, "S") + +lig_dn = paste0("Ligand distance (", angstroms_symbol, ")"); lig_dn +duet_dn = paste0("DUET ", stability_suffix); duet_dn +foldx_dn = paste0("FoldX ", stability_suffix); foldx_dn +deepddg_dn = paste0("Deepddg " , stability_suffix); deepddg_dn +mcsm_na_dn = paste0("mCSM-NA affinity ", stability_suffix); mcsm_na_dn +dynamut_dn = paste0("Dynamut ", stability_suffix); dynamut_dn +dynamut2_dn = paste0("Dynamut2 " , stability_suffix); dynamut2_dn +encom_ddg_dn = paste0("EnCOM " , stability_suffix); encom_ddg_dn +encom_dds_dn = paste0("EnCOM " , flexibility_suffix ); encom_dds_dn +sdm_dn = paste0("SDM " , stability_suffix); sdm_dn +mcsm_dn = paste0("mCSM " , stability_suffix ); mcsm_dn + +# Change colnames of some columns using datatable +comb_df_sl = comb_df_s +names(comb_df_sl) + +setnames(comb_df_sl + , old = c("asa", "rsa", "rd_values", "kd_values" + , "log10_or_mychisq", "neglog_pval_fisher", "af" + , LigDist_colname + , "duet_scaled" + , "foldx_scaled" + , "deepddg_scaled" + , "mcsm_na_scaled" + , "ddg_dynamut_scaled" + , "ddg_dynamut2_scaled" + , "ddg_encom_scaled" + , "dds_encom_scaled" + , "ddg_sdm" + , "ddg_mcsm") + + , new = c("ASA", "RSA", "RD", "KD" + , "Log10 (OR)", "-Log (P)", "MAF" + , lig_dn + , duet_dn + , foldx_dn + , deepddg_dn + , mcsm_na_dn + , dynamut_dn + , dynamut2_dn + , encom_ddg_dn + , encom_dds_dn + , sdm_dn + , mcsm_dn) + ) + +foo_cnames <- cbind(foo_cnames, colnames(comb_df_sl)) + +# some more pretty labels +table(comb_df_sl$mutation_info) + +levels(comb_df_sl$mutation_info)[levels(comb_df_sl$mutation_info)==dr_muts_col] <- "DM" +levels(comb_df_sl$mutation_info)[levels(comb_df_sl$mutation_info)==other_muts_col] <- "OM" + +table(comb_df_sl$mutation_info) + +####################################################################### +#====================== +# Selecting dfs +# with appropriate cols +#======================= +static_cols_start = c("mutationinformation" + , "position" + , "mutation" + , "mutation_info") + +static_cols_end = c(lig_dn + , "ASA" + , "RSA" + , "RD" + , "KD") + +# ordering is important! + +######################################################################### +#============== +# DUET: LF +#============== +cols_to_select_duet = c(static_cols_start, c("duet_outcome", duet_dn), static_cols_end) +wf_duet = comb_df_sl[, cols_to_select_duet] + +#pivot_cols_ps = cols_to_select_ps[1:5]; pivot_cols_ps +pivot_cols_duet = cols_to_select_duet[1: (length(static_cols_start) + 1)]; pivot_cols_duet + +expected_rows_lf = nrow(wf_duet) * (length(wf_duet) - length(pivot_cols_duet)) +expected_rows_lf + +# LF data: duet +lf_duet = gather(wf_duet + , key = param_type + , value = param_value + , all_of(duet_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_duet) == expected_rows_lf){ + cat("\nPASS: long format data created for ", duet_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# FoldX: LF +#============== +cols_to_select_foldx= c(static_cols_start, c("foldx_outcome", foldx_dn), static_cols_end) +wf_foldx = comb_df_sl[, cols_to_select_foldx] + +pivot_cols_foldx = cols_to_select_foldx[1: (length(static_cols_start) + 1)]; pivot_cols_foldx + +expected_rows_lf = nrow(wf_foldx) * (length(wf_foldx) - length(pivot_cols_foldx)) +expected_rows_lf + +# LF data: Foldx +lf_foldx <<- gather(wf_foldx + , key = param_type + , value = param_value + , all_of(foldx_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_foldx) == expected_rows_lf){ + cat("\nPASS: long format data created for ", foldx_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# Deepddg: LF +#============== +cols_to_select_deepddg = c(static_cols_start, c("deepddg_outcome", deepddg_dn), static_cols_end) +wf_deepddg = comb_df_sl[, cols_to_select_deepddg] + +pivot_cols_deepddg = cols_to_select_deepddg[1: (length(static_cols_start) + 1)]; pivot_cols_deepddg + +expected_rows_lf = nrow(wf_deepddg) * (length(wf_deepddg) - length(pivot_cols_deepddg)) +expected_rows_lf + +# LF data: Deepddg +lf_deepddg = gather(wf_deepddg + , key = param_type + , value = param_value + , all_of(deepddg_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_deepddg) == expected_rows_lf){ + cat("\nPASS: long format data created for ", deepddg_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# mCSM-NA: LF +#============== +cols_to_select_mcsm_na = c(static_cols_start, c("mcsm_na_outcome", mcsm_na_dn), static_cols_end) +wf_mcsm_na = comb_df_sl[, cols_to_select_mcsm_na] + +pivot_cols_mcsm_na = cols_to_select_mcsm_na[1: (length(static_cols_start) + 1)]; pivot_cols_mcsm_na + +expected_rows_lf = nrow(wf_mcsm_na) * (length(wf_mcsm_na) - length(pivot_cols_mcsm_na)) +expected_rows_lf + +# LF data: mcsm_na +lf_mcsm_na = gather(wf_mcsm_na + , key = param_type + , value = param_value + , all_of(mcsm_na_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_mcsm_na) == expected_rows_lf){ + cat("\nPASS: long format data created for ", mcsm_na_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# Dynamut: LF +#============== +cols_to_select_dynamut = c(static_cols_start, c("ddg_dynamut_outcome", dynamut_dn), static_cols_end) +wf_dynamut = comb_df_sl[, cols_to_select_dynamut] + +pivot_cols_dynamut = cols_to_select_dynamut[1: (length(static_cols_start) + 1)]; pivot_cols_dynamut + +expected_rows_lf = nrow(wf_dynamut) * (length(wf_dynamut) - length(pivot_cols_dynamut)) +expected_rows_lf + +# LF data: dynamut +lf_dynamut = gather(wf_dynamut + , key = param_type + , value = param_value + , all_of(dynamut_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_dynamut) == expected_rows_lf){ + cat("\nPASS: long format data created for ", dynamut_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# Dynamut2: LF +#============== +cols_to_select_dynamut2 = c(static_cols_start, c("ddg_dynamut2_outcome", dynamut2_dn), static_cols_end) + +wf_dynamut2 = comb_df_sl[, cols_to_select_dynamut2] + +pivot_cols_dynamut2 = cols_to_select_dynamut2[1: (length(static_cols_start) + 1)]; pivot_cols_dynamut2 + +expected_rows_lf = nrow(wf_dynamut2) * (length(wf_dynamut2) - length(pivot_cols_dynamut2)) +expected_rows_lf + +# LF data: dynamut2 +lf_dynamut2 = gather(wf_dynamut2 + , key = param_type + , value = param_value + , all_of(dynamut2_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_dynamut2) == expected_rows_lf){ + cat("\nPASS: long format data created for ", dynamut2_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# EnCOM ddg: LF +#============== +cols_to_select_encomddg = c(static_cols_start, c("ddg_encom_outcome", encom_ddg_dn), static_cols_end) +wf_encomddg = comb_df_sl[, cols_to_select_encomddg] + +pivot_cols_encomddg = cols_to_select_encomddg[1: (length(static_cols_start) + 1)]; pivot_cols_encomddg + +expected_rows_lf = nrow(wf_encomddg ) * (length(wf_encomddg ) - length(pivot_cols_encomddg)) +expected_rows_lf + +# LF data: encomddg +lf_encomddg = gather(wf_encomddg + , key = param_type + , value = param_value + , all_of(encom_ddg_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_encomddg) == expected_rows_lf){ + cat("\nPASS: long format data created for ", encom_ddg_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} +############################################################################ +#============== +# EnCOM dds: LF +#============== +cols_to_select_encomdds = c(static_cols_start, c("dds_encom_outcome", encom_dds_dn), static_cols_end) +wf_encomdds = comb_df_sl[, cols_to_select_encomdds] + +pivot_cols_encomdds = cols_to_select_encomdds[1: (length(static_cols_start) + 1)]; pivot_cols_encomdds + +expected_rows_lf = nrow(wf_encomdds) * (length(wf_encomdds) - length(pivot_cols_encomdds)) +expected_rows_lf + +# LF data: encomdds +lf_encomdds = gather(wf_encomdds + , key = param_type + , value = param_value + , all_of(encom_dds_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_encomdds) == expected_rows_lf){ + cat("\nPASS: long format data created for", encom_dds_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# SDM: LF +#============== +cols_to_select_sdm = c(static_cols_start, c("ddg_sdm_outcome", sdm_dn), static_cols_end) +wf_sdm = comb_df_sl[, cols_to_select_sdm] + +pivot_cols_sdm = cols_to_select_sdm[1: (length(static_cols_start) + 1)]; pivot_cols_sdm + +expected_rows_lf = nrow(wf_sdm) * (length(wf_sdm) - length(pivot_cols_sdm)) +expected_rows_lf + +# LF data: sdm +lf_sdm = gather(wf_sdm + , key = param_type + , value = param_value + , all_of(sdm_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_sdm) == expected_rows_lf){ + cat("\nPASS: long format data created for", sdm_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# mCSM: LF +#============== +cols_to_select_mcsm = c(static_cols_start, c("ddg_mcsm_outcome", mcsm_dn), static_cols_end) +wf_mcsm = comb_df_sl[, cols_to_select_mcsm] + +pivot_cols_mcsm = cols_to_select_mcsm[1: (length(static_cols_start) + 1)]; pivot_cols_mcsm + +expected_rows_lf = nrow(wf_mcsm) * (length(wf_mcsm) - length(pivot_cols_mcsm)) +expected_rows_lf + +# LF data: mcsm +lf_mcsm = gather(wf_mcsm + , key = param_type + , value = param_value + , all_of(mcsm_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_mcsm) == expected_rows_lf){ + cat("\nPASS: long format data created for", mcsm_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +#========================== +# Duet-d(from Dynamut): LF +#=========================== + +#Not created, redundant and chaos! + +############################################################################ + diff --git a/scripts/plotting/logo_data.R b/scripts/plotting/logo_data.R new file mode 100644 index 0000000..7eaf1b6 --- /dev/null +++ b/scripts/plotting/logo_data.R @@ -0,0 +1,142 @@ +#!/usr/bin/env Rscript +######################################################### +# TASK: Script to format data for Logo_plots +######################################################### +#------------------------- +# choose df for logoplot +#------------------------- +logo_data = merged_df3 +#logo_data = merged_df3_comp + +# quick checks +colnames(logo_data) +str(logo_data) + +c1 = unique(logo_data$position) +nrow(logo_data) +cat("No. of rows in my_data:", nrow(logo_data) + , "\nDistinct positions corresponding to snps:", length(c1) + , "\n===========================================================") +#======================================================================= +#================== +# logo data: OR +#================== +foo = logo_data[, c("position" + , "mutant_type","duet_scaled", "or_mychisq" + , "mut_prop_polarity", "mut_prop_water")] + +logo_data$log10or = log10(logo_data$or_mychisq) +logo_data_plot = logo_data[, c("position" + , "mutant_type", "or_mychisq", "log10or")] + +logo_data_plot_or = logo_data[, c("position", "mutant_type", "or_mychisq")] +wide_df_or = logo_data_plot_or %>% spread(position, or_mychisq, fill = 0.0) + +wide_df_or = as.matrix(wide_df_or) +rownames(wide_df_or) = wide_df_or[,1] +dim(wide_df_or) +wide_df_or = wide_df_or[,-1] +str(wide_df_or) + +position_or = as.numeric(colnames(wide_df_or)) + +#================== +# logo data: logOR +#================== +logo_data_plot_logor = logo_data[, c("position", "mutant_type", "log10or")] +wide_df_logor <- logo_data_plot_logor %>% spread(position, log10or, fill = 0.0) + +wide_df_logor = as.matrix(wide_df_logor) + +rownames(wide_df_logor) = wide_df_logor[,1] +wide_df_logor = subset(wide_df_logor, select = -c(1) ) +colnames(wide_df_logor) +wide_df_logor_m = data.matrix(wide_df_logor) + +rownames(wide_df_logor_m) +colnames(wide_df_logor_m) + +position_logor = as.numeric(colnames(wide_df_logor_m)) + +#=============================== +# logo data: multiple nsSNPs (>1) +#================================= +#require(data.table) + +# get freq count of positions so you can subset freq<1 +setDT(logo_data)[, mut_pos_occurrence := .N, by = .(position)] + +table(logo_data$position) +table(logo_data$mut_pos_occurrence) + +max_mut = max(table(logo_data$position)) + +# extract freq_pos > 1 +my_data_snp = logo_data[logo_data$mut_pos_occurrence!=1,] +u = unique(my_data_snp$position) +max_mult_mut = max(table(my_data_snp$position)) + +if (nrow(my_data_snp) == nrow(logo_data) - table(logo_data$mut_pos_occurrence)[[1]] ){ + + cat("PASS: positions with multiple muts extracted" + , "\nNo. of mutations:", nrow(my_data_snp) + , "\nNo. of positions:", length(u) + , "\nMax no. of muts at any position", max_mult_mut) +}else{ + cat("FAIL: positions with multiple muts could NOT be extracted" + , "\nExpected:",nrow(logo_data) - table(logo_data$mut_pos_occurrence)[[1]] + , "\nGot:", nrow(my_data_snp) ) +} + +cat("\nNo. of sites with only 1 mutations:", table(logo_data$mut_pos_occurrence)[[1]]) + +#-------------------------------------- +# matrix for_mychisq mutant type +# frequency of mutant type by position +#--------------------------------------- +table(my_data_snp$mutant_type, my_data_snp$position) +tab_mt = table(my_data_snp$mutant_type, my_data_snp$position) +class(tab_mt) + +# unclass to convert to matrix +tab_mt = unclass(tab_mt) +tab_mt = as.matrix(tab_mt, rownames = T) + +# should be TRUE +is.matrix(tab_mt) + +rownames(tab_mt) #aa +colnames(tab_mt) #pos + +#------------------------------------- +# matrix for wild type +# frequency of wild type by position +#------------------------------------- +tab_wt = table(my_data_snp$wild_type, my_data_snp$position); tab_wt +tab_wt = unclass(tab_wt) + +# remove wt duplicates +wt = my_data_snp[, c("position", "wild_type")] +wt = wt[!duplicated(wt),] + +tab_wt = table(wt$wild_type, wt$position); tab_wt # should all be 1 + +rownames(tab_wt) +rownames(tab_wt) + +identical(colnames(tab_mt), colnames(tab_wt)) +identical(ncol(tab_mt), ncol(tab_wt)) + +#---------------------------------- +# logo data OR: multiple nsSNPs (>1) +#---------------------------------- +logo_data_or_mult = my_data_snp[, c("position", "mutant_type", "or_mychisq")] +#wide_df_or = logo_data_or %>% spread(position, or_mychisq, fill = 0.0) +wide_df_or_mult = logo_data_or_mult %>% spread(position, or_mychisq, fill = NA) + +wide_df_or_mult = as.matrix(wide_df_or_mult) +rownames(wide_df_or_mult) = wide_df_or_mult[,1] +wide_df_or_mult = wide_df_or_mult[,-1] +str(wide_df_or_mult) + +position_or_mult = as.numeric(colnames(wide_df_or_mult)) diff --git a/scripts/plotting/redundant/other_dfs_data.R b/scripts/plotting/redundant/other_dfs_data.R new file mode 100644 index 0000000..97b0567 --- /dev/null +++ b/scripts/plotting/redundant/other_dfs_data.R @@ -0,0 +1,117 @@ +#!/usr/bin/env Rscript + +# Didn't end up using it: sorted it at the source +# .py script to combine all dfs to output all_params + +################################################################# +# TASK: Script to add all other dfs to merged_df2 and merged_df3 + +################################################################# +# Combine other dfs: +# dynamut_df, dynamut2_df, mcsm_na_df, +# perhaps : deepddg and mcsm ppi (for embb) +################################################################ +# read other files +infilename_dynamut = paste0("~/git/Data/", drug, "/output/dynamut_results/", gene + , "_complex_dynamut_norm.csv") + +infilename_dynamut2 = paste0("~/git/Data/", drug, "/output/dynamut_results/dynamut2/", gene + , "_complex_dynamut2_norm.csv") + +infilename_mcsm_na = paste0("~/git/Data/", drug, "/output/mcsm_na_results/", gene + , "_complex_mcsm_na_norm.csv") + +infilename_mcsm_f_snps <- paste0("~/git/Data/", drug, "/output/", gene + , "_mcsm_formatted_snps.csv") + +dynamut_df = read.csv(infilename_dynamut) +dynamut2_df = read.csv(infilename_dynamut2) +mcsm_na_df = read.csv(infilename_mcsm_na) +mcsm_f_snps = read.csv(infilename_mcsm_f_snps, header = F) +names(mcsm_f_snps) = "mutationinformation" + +#================================= +# check with intersect to find the common col, but use +c1 = length(intersect(names(dynamut_df), names(dynamut2_df))) +c2 = length(intersect(names(dynamut2_df), names(mcsm_na_df))) + +if (c1 == 1 && c2 == 1) { + n_common = 1 +}else{ + cat("\nMore than one common col found, inspect before merging!") +} + +# mutationinformation column to be on the safe side +# delete chain from dynamut2_df +#dynamut2_df = subset(dynamut2_df, select = -chain) + +# quick checks +lapply(list(dynamut_df + , dynamut2_df + , mcsm_na_df), ncol) + +lapply(list(dynamut_df + , dynamut2_df + , mcsm_na_df), colnames) + +lapply(list(dynamut_df + , dynamut2_df + , mcsm_na_df), nrow) + +ncols_comb = lapply(list(dynamut_df + , dynamut2_df + , mcsm_na_df), ncol) + +#--------------------------------- +# Combine 1: all other params dfs +#--------------------------------- +combined_dfs = Reduce(inner_join, list(dynamut_df + , dynamut2_df + , mcsm_na_df)) +# Reduce("+", ncols_comb) + +#----------------------------------------- +# Combine 2: combine1 result + merged_df2 +#----------------------------------------- +drop_cols = intersect(names(combined_dfs), names(merged_df2)) +drop_cols = drop_cols + +drop_cols = drop_cols[! drop_cols %in% c("mutationinformation")] + +combined_dfs_f = combined_dfs[, !colnames(combined_dfs)%in%drop_cols] + +nrow(combined_dfs_f); nrow(merged_df2) +ncol(combined_dfs_f); ncol(merged_df2) + +#----------------------------------------- +# Combined merged_df2 +#----------------------------------------- +merged_df2_combined = merge(merged_df2 + , combined_dfs_f + , by = "mutationinformation" +) + +expected_ncols = ncol(combined_dfs_f)+ ncol(merged_df2) - 1 + +if ( nrow(merged_df2_combined) == nrow(merged_df2) && ncol(merged_df2_combined) == expected_ncols ){ + + cat("\nPASS: merged_df2 combined with other parameters dfs." + , "\nUse this for lineage distribution plots") +}else{ + + cat("\nFAIL: merged_df2 didn't combine successfully with other parameters dfs") + quit() + +} + +rm(combined_dfs, combined_dfs_f) + +#================================ +# combined data +# short_df ps: ~ merged_df3 +# TODO: later integrate properly +#================================ +#----------------------------------------- +# Combined merged_df2 +#----------------------------------------- +merged_df3_combined = merged_df2_combined[!duplicated(merged_df2_combined$mutationinformation),] diff --git a/scripts/plotting/redundant/other_plots_data.R b/scripts/plotting/redundant/other_plots_data.R new file mode 100755 index 0000000..61a508f --- /dev/null +++ b/scripts/plotting/redundant/other_plots_data.R @@ -0,0 +1,470 @@ +#!/usr/bin/env Rscript +######################################################### +# TASK: Script to format data for dm om plots: +# generating LF data +# sourced by get_plotting_dfs.R +######################################################### +# working dir and loading libraries +# getwd() +# setwd("~/git/LSHTM_analysis/scripts/plotting") +# getwd() + +# make cmd +# globals +# drug = "streptomycin" +# gene = "gid" + +# source("get_plotting_dfs.R") +#======================================================================= +# MOVE TO COMBINE or singular file for deepddg +# +# cols_to_select = c("mutation", "mutationinformation" +# , "wild_type", "position", "mutant_type" +# , "mutation_info") +# +# merged_df3_short = merged_df3[, cols_to_select] + +# infilename_mcsm_f_snps <- paste0("~/git/Data/", drug, "/output/", gene +# , "_mcsm_formatted_snps.csv") +# +# mcsm_f_snps<- read.csv(infilename_mcsm_f_snps, header = F) +# names(mcsm_f_snps) <- "mutationinformation" + +# write merged_df3 to generate structural figure on chimera +#write.csv(merged_df3_short, "merged_df3_short.csv") +#======================================================================== + +#======================================================================== +# cols to select + +cols_mcsm_df <- merged_df3[, c("mutationinformation", "mutation" + , "mutation_info", "position" + , LigDist_colname + , "duet_stability_change", "duet_scaled", "duet_outcome" + , "ligand_affinity_change", "affinity_scaled", "ligand_outcome" + , "ddg_foldx", "foldx_scaled", "foldx_outcome" + , "deepddg", "deepddg_scaled", "deepddg_outcome" + , "asa", "rsa" + , "rd_values", "kd_values" + , "log10_or_mychisq", "neglog_pval_fisher", "af")] + +cols_mcsm_na_df <- mcsm_na_df[, c("mutationinformation" + , "mcsm_na_affinity", "mcsm_na_scaled" + , "mcsm_na_outcome")] +# entire dynamut_df + +cols_dynamut2_df <- dynamut2_df[, c("mutationinformation" + , "ddg_dynamut2", "ddg_dynamut2_scaled" + , "ddg_dynamut2_outcome")] + +n_comb_cols = length(cols_mcsm_df) + length(cols_mcsm_na_df) + + length(dynamut_df) + length(cols_dynamut2_df); n_comb_cols + +i1<- intersect(names(cols_mcsm_df), names(cols_mcsm_na_df)) +i2<- intersect(names(dynamut_df), names(cols_dynamut2_df)) +merging_cols <- intersect(i1, i2) +cat("\nmerging_cols:", merging_cols) + +if (merging_cols == "mutationinformation") { + cat("\nStage 1: Found common col between dfs, checking values in it...") + c1 <- all(mcsm_f_snps[[merging_cols]]%in%cols_mcsm_df[[merging_cols]]) + c2 <- all(mcsm_f_snps[[merging_cols]]%in%cols_mcsm_na_df[[merging_cols]]) + c3 <- all(mcsm_f_snps[[merging_cols]]%in%dynamut_df[[merging_cols]]) + c4 <- all(mcsm_f_snps[[merging_cols]]%in%cols_dynamut2_df[[merging_cols]]) + cols_check <- c(c1, c2, c3, c4) + expected_cols = n_comb_cols - ( length(cols_check) - 1) + if (all(cols_check)){ + cat("\nStage 2: Proceeding with merging dfs:\n") + comb_df <- Reduce(inner_join, list(cols_mcsm_df + , cols_mcsm_na_df + , dynamut_df + , cols_dynamut2_df)) + comb_df_s = arrange(comb_df, position) + + # if ( nrow(comb_df_s) == nrow(mcsm_f_snps) && ncol(comb_df_s) == expected_cols) { + # cat("\Stage3, PASS: dfs merged sucessfully" + # , "\nnrow of merged_df: ", nrow(comb_df_s) + # , "\nncol of merged_df:", ncol(comb_df_s)) + # } + + } +} +#names(comb_df_s) +cat("\n!!!IT GOT TO HERE!!!!") +#======================================================================= +fact_cols = colnames(comb_df_s)[grepl( "_outcome|_info", colnames(comb_df_s) )] +fact_cols +lapply(comb_df_s[, fact_cols], class) +comb_df_s[, fact_cols] <- lapply(comb_df_s[, fact_cols], as.factor) + +if (any(lapply(comb_df_s[, fact_cols], class) == "character")){ + cat("\nChanging cols to factor") + comb_df_s[, fact_cols] <- lapply(comb_df_s[, fact_cols],as.factor) + if (all(lapply(comb_df_s[, fact_cols], class) == "factor")){ + cat("\nSuccessful: cols changed to factor") + } +} +lapply(comb_df_s[, fact_cols], class) + +#======================================================================= +table(comb_df_s$mutation_info) + + # further checks to make sure dr and other muts are indeed unique +dr_muts = comb_df_s[comb_df_s$mutation_info == dr_muts_col,] +dr_muts_names = unique(dr_muts$mutation) + +other_muts = comb_df_s[comb_df_s$mutation_info == other_muts_col,] +other_muts_names = unique(other_muts$mutation) + +if ( table(dr_muts_names%in%other_muts_names)[[1]] == length(dr_muts_names) && + table(other_muts_names%in%dr_muts_names)[[1]] == length(other_muts_names) ){ + cat("PASS: dr and other muts are indeed unique") +}else{ + cat("FAIL: dr and others muts are NOT unique!") + quit() +} + +# pretty display names i.e. labels to reduce major code duplication later +foo_cnames = data.frame(colnames(comb_df_s)) +names(foo_cnames) <- "old_name" + +stability_suffix <- paste0(delta_symbol, delta_symbol, "G") +flexibility_suffix <- paste0(delta_symbol, delta_symbol, "S") + +lig_dn = paste0("Ligand distance (", angstroms_symbol, ")"); lig_dn +duet_dn = paste0("DUET ", stability_suffix); duet_dn +foldx_dn = paste0("FoldX ", stability_suffix); foldx_dn +deepddg_dn = paste0("Deepddg " , stability_suffix); deepddg_dn +mcsm_na_dn = paste0("mCSM-NA affinity ", stability_suffix); mcsm_na_dn +dynamut_dn = paste0("Dynamut ", stability_suffix); dynamut_dn +dynamut2_dn = paste0("Dynamut2 " , stability_suffix); dynamut2_dn +encom_ddg_dn = paste0("EnCOM " , stability_suffix); encom_ddg_dn +encom_dds_dn = paste0("EnCOM " , flexibility_suffix ); encom_dds_dn +sdm_dn = paste0("SDM " , stability_suffix); sdm_dn +mcsm_dn = paste0("mCSM " , stability_suffix ); mcsm_dn + +# Change colnames of some columns using datatable +comb_df_sl = comb_df_s +names(comb_df_sl) + +setnames(comb_df_sl + , old = c("asa", "rsa", "rd_values", "kd_values" + , "log10_or_mychisq", "neglog_pval_fisher", "af" + , LigDist_colname + , "duet_scaled" + , "foldx_scaled" + , "deepddg_scaled" + , "mcsm_na_scaled" + , "ddg_dynamut_scaled" + , "ddg_dynamut2_scaled" + , "ddg_encom_scaled" + , "dds_encom_scaled" + , "ddg_sdm" + , "ddg_mcsm") + + , new = c("ASA", "RSA", "RD", "KD" + , "Log10 (OR)", "-Log (P)", "MAF" + , lig_dn + , duet_dn + , foldx_dn + , deepddg_dn + , mcsm_na_dn + , dynamut_dn + , dynamut2_dn + , encom_ddg_dn + , encom_dds_dn + , sdm_dn + , mcsm_dn) + ) + +foo_cnames <- cbind(foo_cnames, colnames(comb_df_sl)) + +# some more pretty labels +table(comb_df_sl$mutation_info) + +levels(comb_df_sl$mutation_info)[levels(comb_df_sl$mutation_info)==dr_muts_col] <- "DM" +levels(comb_df_sl$mutation_info)[levels(comb_df_sl$mutation_info)==other_muts_col] <- "OM" + +table(comb_df_sl$mutation_info) + +####################################################################### +#====================== +# Selecting dfs +# with appropriate cols +#======================= +static_cols_start = c("mutationinformation" + , "position" + , "mutation" + , "mutation_info") + +static_cols_end = c(lig_dn + , "ASA" + , "RSA" + , "RD" + , "KD") + +# ordering is important! + +######################################################################### +#============== +# DUET: LF +#============== +cols_to_select_duet = c(static_cols_start, c("duet_outcome", duet_dn), static_cols_end) +wf_duet = comb_df_sl[, cols_to_select_duet] + +#pivot_cols_ps = cols_to_select_ps[1:5]; pivot_cols_ps +pivot_cols_duet = cols_to_select_duet[1: (length(static_cols_start) + 1)]; pivot_cols_duet + +expected_rows_lf = nrow(wf_duet) * (length(wf_duet) - length(pivot_cols_duet)) +expected_rows_lf + +# LF data: duet +lf_duet = gather(wf_duet + , key = param_type + , value = param_value + , all_of(duet_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_duet) == expected_rows_lf){ + cat("\nPASS: long format data created for ", duet_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# FoldX: LF +#============== +cols_to_select_foldx= c(static_cols_start, c("foldx_outcome", foldx_dn), static_cols_end) +wf_foldx = comb_df_sl[, cols_to_select_foldx] + +pivot_cols_foldx = cols_to_select_foldx[1: (length(static_cols_start) + 1)]; pivot_cols_foldx + +expected_rows_lf = nrow(wf_foldx) * (length(wf_foldx) - length(pivot_cols_foldx)) +expected_rows_lf + +# LF data: duet +print("TESTXXXXXXXXXXXXXXXXXXXXX---------------------->>>>") +lf_foldx <<- gather(wf_foldx + , key = param_type + , value = param_value + , all_of(foldx_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_foldx) == expected_rows_lf){ + cat("\nPASS: long format data created for ", foldx_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# Deepddg: LF +#============== +cols_to_select_deepddg = c(static_cols_start, c("deepddg_outcome", deepddg_dn), static_cols_end) +wf_deepddg = comb_df_sl[, cols_to_select_deepddg] + +pivot_cols_deepddg = cols_to_select_deepddg[1: (length(static_cols_start) + 1)]; pivot_cols_deepddg + +expected_rows_lf = nrow(wf_deepddg) * (length(wf_deepddg) - length(pivot_cols_deepddg)) +expected_rows_lf + +# LF data: duet +lf_deepddg = gather(wf_deepddg + , key = param_type + , value = param_value + , all_of(deepddg_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_deepddg) == expected_rows_lf){ + cat("\nPASS: long format data created for ", deepddg_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# mCSM-NA: LF +#============== +cols_to_select_mcsm_na = c(static_cols_start, c("mcsm_na_outcome", mcsm_na_dn), static_cols_end) +wf_mcsm_na = comb_df_sl[, cols_to_select_mcsm_na] + +pivot_cols_mcsm_na = cols_to_select_mcsm_na[1: (length(static_cols_start) + 1)]; pivot_cols_mcsm_na + +expected_rows_lf = nrow(wf_mcsm_na) * (length(wf_mcsm_na) - length(pivot_cols_mcsm_na)) +expected_rows_lf + +# LF data: duet +lf_mcsm_na = gather(wf_mcsm_na + , key = param_type + , value = param_value + , all_of(mcsm_na_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_mcsm_na) == expected_rows_lf){ + cat("\nPASS: long format data created for ", mcsm_na_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# Dynamut: LF +#============== +cols_to_select_dynamut = c(static_cols_start, c("ddg_dynamut_outcome", dynamut_dn), static_cols_end) +wf_dynamut = comb_df_sl[, cols_to_select_dynamut] + +pivot_cols_dynamut = cols_to_select_dynamut[1: (length(static_cols_start) + 1)]; pivot_cols_dynamut + +expected_rows_lf = nrow(wf_dynamut) * (length(wf_dynamut) - length(pivot_cols_dynamut)) +expected_rows_lf + +# LF data: duet +lf_dynamut = gather(wf_dynamut + , key = param_type + , value = param_value + , all_of(dynamut_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_dynamut) == expected_rows_lf){ + cat("\nPASS: long format data created for ", dynamut_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# Dynamut2: LF +#============== +cols_to_select_dynamut2 = c(static_cols_start, c("ddg_dynamut2_outcome", dynamut2_dn), static_cols_end) + +wf_dynamut2 = comb_df_sl[, cols_to_select_dynamut2] + +pivot_cols_dynamut2 = cols_to_select_dynamut2[1: (length(static_cols_start) + 1)]; pivot_cols_dynamut2 + +expected_rows_lf = nrow(wf_dynamut2) * (length(wf_dynamut2) - length(pivot_cols_dynamut2)) +expected_rows_lf + +# LF data: duet +lf_dynamut2 = gather(wf_dynamut2 + , key = param_type + , value = param_value + , all_of(dynamut2_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_dynamut2) == expected_rows_lf){ + cat("\nPASS: long format data created for ", dynamut2_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# EnCOM ddg: LF +#============== +cols_to_select_encomddg = c(static_cols_start, c("ddg_encom_outcome", encom_ddg_dn), static_cols_end) +wf_encomddg = comb_df_sl[, cols_to_select_encomddg] + +pivot_cols_encomddg = cols_to_select_encomddg[1: (length(static_cols_start) + 1)]; pivot_cols_encomddg + +expected_rows_lf = nrow(wf_encomddg ) * (length(wf_encomddg ) - length(pivot_cols_encomddg)) +expected_rows_lf + +# LF data: encomddg +lf_encomddg = gather(wf_encomddg + , key = param_type + , value = param_value + , all_of(encom_ddg_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_encomddg) == expected_rows_lf){ + cat("\nPASS: long format data created for ", encom_ddg_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} +############################################################################ +#============== +# EnCOM dds: LF +#============== +cols_to_select_encomdds = c(static_cols_start, c("dds_encom_outcome", encom_dds_dn), static_cols_end) +wf_encomdds = comb_df_sl[, cols_to_select_encomdds] + +pivot_cols_encomdds = cols_to_select_encomdds[1: (length(static_cols_start) + 1)]; pivot_cols_encomdds + +expected_rows_lf = nrow(wf_encomdds) * (length(wf_encomdds) - length(pivot_cols_encomdds)) +expected_rows_lf + +# LF data: encomddg +lf_encomdds = gather(wf_encomdds + , key = param_type + , value = param_value + , all_of(encom_dds_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_encomdds) == expected_rows_lf){ + cat("\nPASS: long format data created for", encom_dds_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# SDM: LF +#============== +cols_to_select_sdm = c(static_cols_start, c("ddg_sdm_outcome", sdm_dn), static_cols_end) +wf_sdm = comb_df_sl[, cols_to_select_sdm] + +pivot_cols_sdm = cols_to_select_sdm[1: (length(static_cols_start) + 1)]; pivot_cols_sdm + +expected_rows_lf = nrow(wf_sdm) * (length(wf_sdm) - length(pivot_cols_sdm)) +expected_rows_lf + +# LF data: encomddg +lf_sdm = gather(wf_sdm + , key = param_type + , value = param_value + , all_of(sdm_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_sdm) == expected_rows_lf){ + cat("\nPASS: long format data created for", sdm_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +############################################################################ +#============== +# mCSM: LF +#============== +cols_to_select_mcsm = c(static_cols_start, c("ddg_mcsm_outcome", mcsm_dn), static_cols_end) +wf_mcsm = comb_df_sl[, cols_to_select_mcsm] + +pivot_cols_mcsm = cols_to_select_mcsm[1: (length(static_cols_start) + 1)]; pivot_cols_mcsm + +expected_rows_lf = nrow(wf_mcsm) * (length(wf_mcsm) - length(pivot_cols_mcsm)) +expected_rows_lf + +# LF data: encomddg +lf_mcsm = gather(wf_mcsm + , key = param_type + , value = param_value + , all_of(mcsm_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_mcsm) == expected_rows_lf){ + cat("\nPASS: long format data created for", mcsm_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} +############################################################################ +