diff --git a/scripts/functions/dm_om_data.R b/scripts/functions/dm_om_data.R new file mode 100644 index 0000000..221e0fc --- /dev/null +++ b/scripts/functions/dm_om_data.R @@ -0,0 +1,710 @@ +#!/usr/bin/env Rscript +######################################################### +# TASK: Script to format data for dm om plots: + # generating WF and LF data for each of the parameters + # duet, mcsm-lig, foldx, deepddg, dynamut2, mcsm-na, mcsm-ppi2, encom, dynamut..etc + # Called by get_plotting_dfs.R + +# dm_om_wf_lf_data() +# Input: data with all parameters (merged_df3, my_use case) +# gene: [conditional generation of dfs like mcsm-NA, mcsm-ppi2 as not all genes have all these values] +# colnames_to_extract = c("mutationinformation" +# , "duet_affinity_change...") +# ligand_dist_colname = LigDist_colname # from globals +# dr_muts = dr_muts_col # from globals ...dr_mutations_ +# other_muts = other_muts_col # from globals ...other_mutations_ +# snp_colname = "mutationinformation" +# aa_pos_colname = "position" # to sort df by +# mut_colname = "mutation" +# mut_info_colname = "mutation_info" +# mut_info_label_colname = "mutation_info_labels" # if empty, below used +# dr_other_muts_labels = c("DM", "OM") # only used if ^^ = "" +# categ_cols_to_factor: converts the cols with '_outcome'and 'info' to factor + +# TO DO: SHINY +#1) +#2) +################################################################## +dm_om_wf_lf_data <- function(df + , gene_name = gene # from globals + , colnames_to_extract + , ligand_dist_colname = LigDist_colname # from globals + , dr_muts = dr_muts_col # from globals + , other_muts = other_muts_col # from globals + , snp_colname = "mutationinformation" + , aa_pos_colname = "position" # to sort df by + , mut_colname = "mutation" + , mut_info_colname = "mutation_info" + , mut_info_label_colname = "mutation_info_labels" # if empty, below used + , dr_other_muts_labels = c("DM", "OM") # only used if ^^ = "" + , categ_cols_to_factor){ + + # Initialise the required dfs based on gene name + geneL_normal = c("pnca") + #geneL_na_dy = c("gid") + geneL_na = c("gid", "rpob") + geneL_dy = c("gid") + geneL_ppi2 = c("alr", "embb", "katg", "rpob") + + # common_dfs + common_dfsL = list( + wf_duet = data.frame() + , lf_duet = data.frame() + , wf_mcsm_lig = data.frame() + , lf_mcsm_lig = data.frame() + , wf_foldx = data.frame() + , lf_foldx = data.frame() + , wf_deepddg = data.frame() + , lf_deepddg = data.frame() + , wf_dynamut2 = data.frame() + , lf_dynamut2 = data.frame() + , wf_consurf = data.frame() + , lf_consurf = data.frame() + , wf_snap2 = data.frame() + , lf_snap2 = data.frame() + ) + + # additional dfs + if (tolower(gene_name)%in%geneL_normal){ + wf_lf_dataL = common_dfsL + } + + if (tolower(gene_name)%in%geneL_na){ + additional_dfL = list( + wf_mcsm_na = data.frame() + , lf_mcsm_na = data.frame() + ) + wf_lf_dataL = c(common_dfsL, additional_dfL) + } + + if (tolower(gene_name)%in%geneL_ppi2){ + additional_dfL = list( + wf_mcsm_ppi2 = data.frame() + , lf_mcsm_ppi2 = data.frame() + ) + wf_lf_dataL = c(common_dfsL, additional_dfL) + } + + if (tolower(gene_name)%in%geneL_dy){ + additional_dfL = list( + wf_mcsm_na = data.frame() + , lf_mcsm_na = data.frame() + , wf_dynamut = data.frame() + , lf_dynamut = data.frame() + , wf_encomddg = data.frame() + , lf_encomddg = data.frame() + , wf_encomdds = data.frame() + , lf_encomdds = data.frame() + , wf_sdm = data.frame() + , lf_sdm = data.frame() + , wf_mcsm = data.frame() + , lf_mcsm = data.frame() + ) + wf_lf_dataL = c(common_dfsL, additional_dfL) + } + cat("\nInitializing an empty list of length:" + , length(wf_lf_dataL)) + + #======================================================================= + if (missing(colnames_to_extract)){ + + colnames_to_extract = c(snp_colname + , mut_colname, mut_info_colname, mut_info_label_colname + , aa_pos_colname + , 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" + , "ddg_dynamut2" , "ddg_dynamut2_scaled", "ddg_dynamut2_outcome" + , "mcsm_ppi2_affinity" , "mcsm_ppi2_scaled" , "mcsm_ppi2_outcome" + , "consurf_score" , "consurf_scaled" #, "consurf_outcome" + , "snap2_score" , "snap2_scaled" , "snap2_outcome" + , "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") + }else{ + colnames_to_extract = c(mut_colname, mut_info_colname, mut_info_label_colname + , aa_pos_colname, LigDist_colname + , colnames_to_extract) + } + comb_df = df[, colnames(df)%in%colnames_to_extract] + comb_df_s = dplyr::arrange(comb_df, aa_pos_colname) + +#======================================================================= + if(missing(categ_cols_to_factor)){ + categ_cols_to_factor = grep( "_outcome|_info", colnames(comb_df_s) ) + }else{ + categ_cols_to_factor = categ_cols_to_factor + } + #fact_cols = colnames(comb_df_s)[grepl( "_outcome|_info", colnames(comb_df_s) )] + fact_cols = colnames(comb_df_s)[categ_cols_to_factor] + + if (any(lapply(comb_df_s[, fact_cols], class) == "character")){ + cat("\nChanging", length(categ_cols_to_factor), "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") + } + }else{ + cat("\nRequested cols aready factors") + } +#======================================================================= +table(comb_df_s[[mut_info_colname]]) + +# further checks to make sure dr and other muts are indeed unique +dr_muts = comb_df_s[comb_df_s[[mut_info_colname]] == dr_muts,] +dr_muts_names = unique(dr_muts$mutation) + +other_muts = comb_df_s[comb_df_s[[mut_info_colname]] == other_muts,] +other_muts_names = unique(other_muts$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 +mcsm_lig_dn = paste0("Ligand affinity (log fold change)"); mcsm_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 +dynamut2_dn = paste0("Dynamut2 " , stability_suffix); dynamut2_dn + +mcsm_na_dn = paste0("mCSM-NA affinity ", stability_suffix); mcsm_na_dn +mcsm_ppi2_dn = paste0("mCSM-PPI2 affinity ", stability_suffix); mcsm_ppi2_dn +consurf_dn = paste0("Consurf"); consurf_dn +snap2_dn = paste0("SNAP2"); snap2_dn + +dynamut_dn = paste0("Dynamut ", stability_suffix); dynamut_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 column names: plyr +new_colnames = c(asa = "ASA" + , rsa = "RSA" + , rd_values = "RD" + , kd_values = "KD" + , log10_or_mychisq = "Log10 (OR)" + , neglog_pval_fisher = "-Log (P)" + , af = "MAF" + #, ligand_dist_colname = lig_dn # cannot handle variable name 'ligand_dist_colname' + , affinity_scaled = mcsm_lig_dn + , duet_scaled = duet_dn + , foldx_scaled = foldx_dn + , deepddg_scaled = deepddg_dn + , ddg_dynamut2_scaled = dynamut2_dn + , mcsm_na_scaled = mcsm_na_dn + , mcsm_ppi2_affinity = mcsm_ppi2_dn + , consurf_score = consurf_dn + , snap2_score = snap2_dn + , ddg_dynamut_scaled = dynamut_dn + , ddg_encom_scaled = encom_ddg_dn + , dds_encom_scaled = encom_dds_dn + , ddg_sdm = sdm_dn + , ddg_mcsm = mcsm_dn) + +comb_df_sl1 = plyr::rename(comb_df_s + , replace = new_colnames + , warn_missing = T + , warn_duplicated = T) + +# renaming colname using variable i.e ligand_dist_colname: dplyr +comb_df_sl = comb_df_sl1 %>% dplyr::rename(!!lig_dn := all_of(ligand_dist_colname)) +names(comb_df_sl) +##################################################################### +if (mut_info_label_colname == "") { + cat("\nAssigning labels:", dr_other_muts_labels, "--> to column:", mut_info_colname) + table(comb_df_sl[[mut_info_colname]]) + + # dr_muts + levels(comb_df_sl[[mut_info_colname]])[levels(comb_df_sl[[mut_info_colname]])==dr_muts] <- dr_other_muts_labels[[1]] + # other_muts + levels(comb_df_sl[[mut_info_colname]])[levels(comb_df_sl[[mut_info_colname]])==other_muts] <- dr_other_muts_labels[[2]] + table(comb_df_sl[[mut_info_colname]]) + + static_cols1 = mut_info_colname +}else{ + table(comb_df_sl[[mut_info_label_colname]]) + static_cols1 = mut_info_label_colname + +} +####################################################################### +#====================== +# Selecting dfs +# with appropriate cols +#======================= + +static_cols_start = c(snp_colname + , aa_pos_colname + , mut_colname + , static_cols1) + +# ordering is important! +static_cols_end = c(lig_dn + , "ASA" + , "RSA" + , "RD" + , "KD" + , "MAF" + , "Log10 (OR)" + , "-Log (P)") + +######################################################################### +#============== +# DUET +#============== +# WF data: duet +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() +} + +# Assign them to the output list +wf_lf_dataL[['wf_duet']] = wf_duet +wf_lf_dataL[['lf_duet']] = lf_duet + +############################################################################ +#============== +# mCSM-lig +#============== +# WF data: mcsm_lig +cols_to_select_mcsm_lig = c(static_cols_start, c("ligand_outcome", mcsm_lig_dn), static_cols_end) +wf_mcsm_lig = comb_df_sl[, cols_to_select_mcsm_lig] + +pivot_cols_mcsm_lig = cols_to_select_mcsm_lig[1: (length(static_cols_start) + 1)]; pivot_cols_mcsm_lig +expected_rows_lf = nrow(wf_mcsm_lig) * (length(wf_mcsm_lig) - length(pivot_cols_mcsm_lig)) +expected_rows_lf + +# LF data: mcsm_lig +lf_mcsm_lig = gather(wf_mcsm_lig + , key = param_type + , value = param_value + , all_of(mcsm_lig_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_mcsm_lig) == expected_rows_lf){ + cat("\nPASS: long format data created for ", mcsm_lig_dn) +}else{ + cat("\nFAIL: long format data could not be created for mcsm_lig") + quit() +} + +# Assign them to the output list +wf_lf_dataL[['wf_mcsm_lig']] = wf_mcsm_lig +wf_lf_dataL[['lf_mcsm_lig']] = lf_mcsm_lig +############################################################################ +#============== +# FoldX +#============== +# WF data: Foldx +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() +} + +# Assign them to the output list +wf_lf_dataL[['wf_foldx']] = wf_foldx +wf_lf_dataL[['lf_foldx']] = lf_foldx + +############################################################################ +#============== +# Deepddg +#============== +# WF data: deepddg +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() +} + +# Assign them to the output list +wf_lf_dataL[['wf_deepddg']] = wf_deepddg +wf_lf_dataL[['lf_deepddg']] = lf_deepddg +############################################################################ +#============== +# Dynamut2: LF +#============== +# WF data: dynamut2 +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() +} + +# Assign them to the output list +wf_lf_dataL[['wf_dynamut2']] = wf_dynamut2 +wf_lf_dataL[['lf_dynamut2']] = lf_dynamut2 +############################################################################ +#================== +# Consurf: LF +#https://consurf.tau.ac.il/overview.php +# consurf_score: +# <0 (below average): slowly evolving i.e CONSERVED +# >0 (above average): rapidly evolving, i.e VARIABLE +#table(df$consurf_colour_rev) +# TODO +#1--> "most_variable", 2--> "", 3-->"", 4-->"" +#5-->"", 6-->"", 7-->"", 8-->"", 9-->"most_conserved" +#==================== +# FIXME: if you add category column to consurf +#cols_to_select_consurf = c(static_cols_start, c("consurf_outcome", consurf_dn), static_cols_end) +#wf_consurf = comb_df_sl[, cols_to_select_consurf] +#pivot_cols_consurf = cols_to_select_consurf[1: (length(static_cols_start) + 1)]; pivot_cols_consurf + +# WF data: consurf +cols_to_select_consurf = c(static_cols_start, c(consurf_dn), static_cols_end) +wf_consurf = comb_df_sl[, cols_to_select_consurf] + +pivot_cols_consurf = cols_to_select_consurf[1: (length(static_cols_start))]; pivot_cols_consurf +expected_rows_lf = nrow(wf_consurf) * (length(wf_consurf) - length(pivot_cols_consurf)) +expected_rows_lf + +# LF data: consurf +lf_consurf = gather(wf_consurf + , key = param_type + , value = param_value + , all_of(consurf_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_consurf) == expected_rows_lf){ + cat("\nPASS: long format data created for ", consurf_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +# Assign them to the output list +wf_lf_dataL[['wf_consurf']] = wf_consurf +wf_lf_dataL[['lf_consurf']] = lf_consurf +########################################################################### +#============== +# SNAP2: LF +#============== +# WF data: snap2 +cols_to_select_snap2 = c(static_cols_start, c("snap2_outcome", snap2_dn), static_cols_end) +wf_snap2 = comb_df_sl[, cols_to_select_snap2] + +pivot_cols_snap2 = cols_to_select_snap2[1: (length(static_cols_start) + 1)]; pivot_cols_snap2 +expected_rows_lf = nrow(wf_snap2) * (length(wf_snap2) - length(pivot_cols_snap2)) +expected_rows_lf + +# LF data: snap2 +lf_snap2 = gather(wf_snap2 + , key = param_type + , value = param_value + , all_of(snap2_dn):tail(static_cols_end,1) + , factor_key = TRUE) + +if (nrow(lf_snap2) == expected_rows_lf){ + cat("\nPASS: long format data created for ", snap2_dn) +}else{ + cat("\nFAIL: long format data could not be created for duet") + quit() +} + +# Assign them to the output list +wf_lf_dataL[['wf_snap2']] = wf_snap2 +wf_lf_dataL[['lf_snap2']] = lf_snap2 + +############################################################################ +if (tolower(gene_name)%in%geneL_na){ + #============== + # mCSM-NA: LF + #============== + # WF data: mcsm-na + 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() + } + + # Assign them to the output list + wf_lf_dataL[['wf_mcsm_na']] = wf_mcsm_na + wf_lf_dataL[['lf_mcsm_na']] = lf_mcsm_na + +} +#------------------------------------------------------------------- +if (tolower(gene_name)%in%geneL_ppi2){ + #============== + # mCSM-PPI2: LF + #============== + # WF data: mcsm-ppi2 + cols_to_select_mcsm_ppi2 = c(static_cols_start, c("mcsm_ppi2_outcome", mcsm_ppi2_dn), static_cols_end) + wf_mcsm_ppi2 = comb_df_sl[, cols_to_select_mcsm_ppi2] + + pivot_cols_mcsm_ppi2 = cols_to_select_mcsm_ppi2[1: (length(static_cols_start) + 1)]; pivot_cols_mcsm_ppi2 + expected_rows_lf = nrow(wf_mcsm_ppi2) * (length(wf_mcsm_ppi2) - length(pivot_cols_mcsm_ppi2)) + expected_rows_lf + + # LF data: mcsm-ppi2 + lf_mcsm_ppi2 = gather(wf_mcsm_ppi2 + , key = param_type + , value = param_value + , all_of(mcsm_ppi2_dn):tail(static_cols_end,1) + , factor_key = TRUE) + + if (nrow(lf_mcsm_ppi2) == expected_rows_lf){ + cat("\nPASS: long format data created for ", mcsm_ppi2_dn) + }else{ + cat("\nFAIL: long format data could not be created for duet") + quit() + } + + # Assign them to the output list + wf_lf_dataL[['wf_mcsm_ppi2']] = wf_mcsm_ppi2 + wf_lf_dataL[['lf_mcsm_ppi2']] = lf_mcsm_ppi2 + +} +#------------------------------------------------------------------- +if (tolower(gene_name)%in%geneL_dy){ + #============== + # Dynamut: LF + #============== + # WF data: dynamut + 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() + } + + # Assign them to the output list + wf_lf_dataL[['wf_dynamut']] = wf_dynamut + wf_lf_dataL[['lf_dynamut']] = lf_dynamut + +#------------------------------------------------------------------------- + #============== + # EnCOM ddg: LF + #============== + # WF data: encomddg + 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() + } + + # Assign them to the output list + wf_lf_dataL[['wf_encomddg']] = wf_encomddg + wf_lf_dataL[['lf_encomddg']] = lf_encomddg +#------------------------------------------------------------------------- + #============== + # EnCOM dds: LF + #============== + # WF data: encomdds + 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() + } + + # Assign them to the output list + wf_lf_dataL[['wf_encomdds']] = wf_encomdds + wf_lf_dataL[['lf_encomdds']] = lf_encomdds +#------------------------------------------------------------------------- + #============== + # SDM: LF + #============== + # WF data: sdm + 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() + } + + # Assign them to the output list + wf_lf_dataL[['wf_sdm']] = wf_sdm + wf_lf_dataL[['lf_sdm']] = lf_sdm +#------------------------------------------------------------------------- + #============== + # mCSM: LF + #============== + # WF data: mcsm + 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() + } + + # Assign them to the output list + wf_lf_dataL[['wf_mcsm']] = wf_mcsm + wf_lf_dataL[['lf_mcsm']] = lf_mcsm + + } +#------------------------------------------------------------------------- +return(wf_lf_dataL) +} +############################################################################ diff --git a/scripts/functions/tests/test_dm_om_data.R b/scripts/functions/tests/test_dm_om_data.R new file mode 100644 index 0000000..d1391cf --- /dev/null +++ b/scripts/functions/tests/test_dm_om_data.R @@ -0,0 +1,53 @@ +#!/usr/bin/env Rscript +#source("~/git/LSHTM_analysis/config/alr.R") +#source("~/git/LSHTM_analysis/config/embb.R") +source("~/git/LSHTM_analysis/config/gid.R") +#source("~/git/LSHTM_analysis/config/katg.R") +#source("~/git/LSHTM_analysis/config/pnca.R") +#source("~/git/LSHTM_analysis/config/rpob.R") + +source("~/git/LSHTM_analysis/scripts/plotting/get_plotting_dfs.R") + +all_dm_om_df = dm_om_wf_lf_data(df = merged_df3, gene_name = gene) + +wf_duet = all_dm_om_df[['wf_duet']] +lf_duet = all_dm_om_df[['lf_duet']] + +wf_mcsm_lig = all_dm_om_df[['wf_mcsm_lig']] +lf_mcsm_lig = all_dm_om_df[['lf_mcsm_lig']] + +wf_foldx = all_dm_om_df[['wf_foldx']] +lf_foldx = all_dm_om_df[['lf_foldx']] + +wf_deepddg = all_dm_om_df[['wf_deepddg']] +lf_deepddg = all_dm_om_df[['lf_deepddg']] + +wf_dynamut2 = all_dm_om_df[['wf_dynamut2']] +lf_dynamut2 = all_dm_om_df[['lf_dynamut2']] + +wf_consurf = all_dm_om_df[['wf_consurf']] +lf_consurf = all_dm_om_df[['lf_consurf']] + +wf_snap2 = all_dm_om_df[['wf_snap2']] +lf_snap2 = all_dm_om_df[['lf_snap2']] + +wf_mcsm_na = all_dm_om_df[['wf_mcsm_na']] +lf_mcsm_na = all_dm_om_df[['lf_mcsm_na']] + +wf_mcsm_ppi2 = all_dm_om_df[['wf_mcsm_ppi2']] +lf_mcsm_ppi2 = all_dm_om_df[['lf_mcsm_ppi2']] + +wf_dynamut = all_dm_om_df[['wf_dynamut']] +lf_dynamut = all_dm_om_df[['lf_dynamut']] + +wf_encomddg = all_dm_om_df[['wf_encomddg']] +lf_encomddg = all_dm_om_df[['lf_encomddg']] + +wf_encomdds = all_dm_om_df[['wf_encomdds']] +lf_encomdds = all_dm_om_df[['lf_encomdds']] + +wf_sdm = all_dm_om_df[['wf_sdm']] +lf_sdm = all_dm_om_df[['lf_sdm']] + +wf_mcsm = all_dm_om_df[['wf_mcsm']] +lf_mcsm = all_dm_om_df[['lf_mcsm']] \ No newline at end of file