tanu/LSHTM_analysis
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

git trimmed downthe dm_om_data.R

Browse source
This commit is contained in:
Tanushree Tunstall 2022-08-05 14:36:02 +01:00
parent fae846395d
commit 05ab89ec09
5 changed files with 168 additions and 351 deletions
Download patch file Download diff file Expand all files Collapse all files

43
scripts/functions/combining_dfs_plotting.R
View file

@ -343,20 +343,45 @@ combining_dfs_plotting <- function( my_df_u
, "\nNo. of rows merged_df3: ", nrow(merged_df3)) , "\nNo. of rows merged_df3: ", nrow(merged_df3))
quit() quit()
} }
#=========================================
# NEW: add consurf outcome
#=========================================
consurf_colOld = "consurf_colour_rev"
consurf_colNew = "consurf_outcome"
merged_df3[[consurf_colNew]] = merged_df3[[consurf_colOld]]
merged_df3[[consurf_colNew]] = as.factor(merged_df3[[consurf_colNew]])
merged_df3[[consurf_colNew]]
#levels(merged_df3$consurf_outcome) = c("nsd", 1, 2, 3, 4, 5, 6, 7, 8, 9)
merged_df2[[consurf_colNew]] = merged_df2[[consurf_colOld]]
merged_df2[[consurf_colNew]] = as.factor(merged_df2[[consurf_colNew]])
merged_df2[[consurf_colNew]]
#=========================================
# NEW: fixed case for SNAP2 labels
#=========================================
snap2_colname = "snap2_outcome"
merged_df3[[snap2_colname]] <- str_replace(merged_df3[[snap2_colname]], "effect", "Effect")
merged_df3[[snap2_colname]] <- str_replace(merged_df3[[snap2_colname]], "neutral", "Neutral")
merged_df2[[snap2_colname]] <- str_replace(merged_df2[[snap2_colname]], "effect", "Effect")
merged_df2[[snap2_colname]] <- str_replace(merged_df2[[snap2_colname]], "neutral", "Neutral")
#--------------------------------------------- #---------------------------------------------
# add columns that are needed to generate plots with revised colnames and strings # NEW: add columns that are needed to generate
# plots with revised colnames and strings
#---------------------------------------------- #----------------------------------------------
merged_df3['sensitivity'] = ifelse(merged_df3['dst_mode'] == 1, "R", "S") merged_df3$sensitivity = ifelse(merged_df3$dst_mode == 1, "R", "S")
merged_df3['mutation_info_labels'] = ifelse(merged_df3['mutation_info_labels'] == "DM", "R", "S") merged_df3$mutation_info_labels = ifelse(merged_df3$mutation_info_labels == "DM", "R", "S")
merged_df2['sensitivity'] = ifelse(merged_df2['dst_mode'] == 1, "R", "S") merged_df2$sensitivity = ifelse(merged_df2$dst_mode == 1, "R", "S")
merged_df2['mutation_info_labels'] = ifelse(merged_df2['mutation_info_labels'] == "DM", "R", "S") merged_df2$mutation_info_labels = ifelse(merged_df2$mutation_info_labels == "DM", "R", "S")
#check1 = all(table(merged_df3["mutation_info_labels"]) == table(merged_df3['sensitivity'])) # for epistasis: fill na where dst: No equivalent in merged_df3
#check2 = all(table(merged_df2["mutation_info_labels"]) == table(merged_df2['sensitivity'])) merged_df2$dst2 = ifelse(is.na(merged_df2$dst), merged_df2$dst_mode, merged_df2$dst)
check1 = all(merged_df3["mutation_info_labels"] == merged_df3['sensitivity']) check1 = all(merged_df3$mutation_info_labels == merged_df3$sensitivity)
check2 = all(merged_df2["mutation_info_labels"] == merged_df2['sensitivity']) check2 = all(merged_df2$mutation_info_labels == merged_df2$sensitivity)
if(check1 && check2){ if(check1 && check2){
cat("PASS: merged_df3 and merged_df2 have mutation info labels as R and S" cat("PASS: merged_df3 and merged_df2 have mutation info labels as R and S"

421
scripts/functions/dm_om_data.R
View file

@ -5,47 +5,16 @@
# duet, mcsm-lig, foldx, deepddg, dynamut2, mcsm-na, mcsm-ppi2, encom, dynamut..etc # duet, mcsm-lig, foldx, deepddg, dynamut2, mcsm-na, mcsm-ppi2, encom, dynamut..etc
# Called by get_plotting_dfs.R # Called by get_plotting_dfs.R
# dm_om_wf_lf_data()
# INPUT:
# df: merged_df3 (data with all parameters)
# NOTE*: merged_df2 will not be appropriate as it brings up most params as significant!?,atleast for gid
# gene: [conditional generation of dfs like mcsm-NA, mcsm-ppi2 as not all genes have all these values]
# colnames_to_extract : columns to extract, either user-specified.
#By default it is c("mutationinformation" , "duet_affinity_change...")
# ligand_dist_colname : column name containing ligand distance. By deafult, it is LigDist_colname (imported from globals)
# dr_muts : dr_muts_col (imported from globals; dr_mutations_<drug>)
# other_muts : other_muts_col (imported from globals ...other_mutations_<drug>)
# snp_colname : SNP column name. By default it is "mutationinformation"
# aa_pos_colname : Column name containing the aa position. This is used to sort the df by.
# mut_colname : Column name containing snp info in format "<abc_pXXdef>. By default, it is "mutation"
# mut_info_colname : Column name containing mutation info whether it is DM or OM. By default, it is "mutation_info"
# mut_info_label_colname : Column containing pre-formatted labels for mutation info.
# For my use case, this is called "mutation_info_labels"
# This column has short labels like DM and OM coresponding to dr_muts and other_muts.
# NOTE*: if this is left empty, then the arg ('dr_other_muts_labels') will be used
# dr_other_muts_labels : User specified labels, must correspond to dr_muts and other_muts.
# NOTE*: Only used if the arg (mut_info_label_colname) is empty!
# categ_cols_to_factor : Column names to convert to factors. These mainly correspond to the outcome columns associated with the
# arg ('colnames_to_extract'). These have the suffix "_outcome" in their colnames. Additionally column 'mutation_info' is also
# converted to factor. By default, it converts the cols with '_outcome'and 'info' to factor.
# Users are able to provide a vector of their corresponding column names
# RETURNS: List
# WF nd LF data grouped by mutation_info i.e DM (drug mutations) and OM (other mutations)
# TO DO: SHINY
#1) df to choose (merged_df3 or merged_df2)
#2)
################################################################## ##################################################################
DistCutOff = 10 # from plotting_globals.R
#LigDist_colname # = "ligand_distance" # from globals # DistCutOff, LigDist_colname, ppi2Dist_colname, naDist_colname
ppi2Dist_colname = "interface_dist"
naDist_colname = "TBC"
dm_om_wf_lf_data <- function(df dm_om_wf_lf_data <- function(df
, gene_name = gene # from globals , gene_name = gene # from globals
, colnames_to_extract , colnames_to_extract
, ligand_dist_colname = LigDist_colname # from globals , ligand_dist_colname = LigDist_colname # from globals
#, ppi2Dist_colname #from globals used
#, naDist_colname #from globals used
, dr_muts = dr_muts_col # from globals , dr_muts = dr_muts_col # from globals
, other_muts = other_muts_col # from globals , other_muts = other_muts_col # from globals
, snp_colname = "mutationinformation" , snp_colname = "mutationinformation"
@ -53,28 +22,19 @@ dm_om_wf_lf_data <- function(df
, mut_colname = "mutation" , mut_colname = "mutation"
, mut_info_colname = "mutation_info" , mut_info_colname = "mutation_info"
, mut_info_label_colname = "mutation_info_labels" # if empty, below used , mut_info_label_colname = "mutation_info_labels" # if empty, below used
, dr_other_muts_labels = c("DM", "OM") # only used if ^^ = "" #, dr_other_muts_labels = c("DM", "OM") # only used if ^^ = ""
, categ_cols_to_factor){ , categ_cols_to_factor){
df = as.data.frame(df) df = as.data.frame(df)
df['sensitivity'] = ifelse(df['dst_mode'] == 1, "R", "S")
table(df['sensitivity'])
df[[mut_info_label_colname]] = ifelse(df[[mut_info_label_colname]] == "DM", "R", "S")
table(df[[mut_info_label_colname]])
# Initialise the required dfs based on gene name # Initialise the required dfs based on gene name
geneL_normal = c("pnca") geneL_normal = c("pnca")
#geneL_na_dy = c("gid")
geneL_na = c("gid", "rpob") geneL_na = c("gid", "rpob")
geneL_dy = c("gid")
geneL_ppi2 = c("alr", "embb", "katg", "rpob") geneL_ppi2 = c("alr", "embb", "katg", "rpob")
# common_dfs # common_dfs
common_dfsL = list( common_dfsL = list(
wf_duet = data.frame() wf_duet = data.frame()
, lf_duet = data.frame() , lf_duet = data.frame()
, wf_mcsm_lig = data.frame() , wf_mcsm_lig = data.frame()
, lf_mcsm_lig = data.frame() , lf_mcsm_lig = data.frame()
@ -110,24 +70,6 @@ dm_om_wf_lf_data <- function(df
) )
wf_lf_dataL = c(common_dfsL, additional_dfL) 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:" cat("\nInitializing an empty list of length:"
, length(wf_lf_dataL)) , length(wf_lf_dataL))
@ -137,26 +79,21 @@ dm_om_wf_lf_data <- function(df
colnames_to_extract = c(snp_colname colnames_to_extract = c(snp_colname
, mut_colname, mut_info_colname, mut_info_label_colname , mut_colname, mut_info_colname, mut_info_label_colname
, aa_pos_colname , aa_pos_colname
, LigDist_colname , LigDist_colname # from globals
, ppi2Dist_colname, naDist_colname , ppi2Dist_colname # from globals
, naDist_colname # from globals
, "duet_stability_change" , "duet_scaled" , "duet_outcome" , "duet_stability_change" , "duet_scaled" , "duet_outcome"
, "ligand_affinity_change", "affinity_scaled" , "ligand_outcome" , "ligand_affinity_change", "affinity_scaled" , "ligand_outcome"
, "ddg_foldx" , "foldx_scaled" , "foldx_outcome" , "ddg_foldx" , "foldx_scaled" , "foldx_outcome"
, "deepddg" , "deepddg_scaled" , "deepddg_outcome" , "deepddg" , "deepddg_scaled" , "deepddg_outcome"
, "asa" , "rsa" , "asa" , "rsa"
, "rd_values" , "kd_values" , "rd_values" , "kd_values"
, "log10_or_mychisq" , "neglog_pval_fisher" , "af" , "log10_or_mychisq" , "neglog_pval_fisher" , "maf" #"af"
, "ddg_dynamut2" , "ddg_dynamut2_scaled", "ddg_dynamut2_outcome" , "ddg_dynamut2" , "ddg_dynamut2_scaled", "ddg_dynamut2_outcome"
, "mcsm_ppi2_affinity" , "mcsm_ppi2_scaled" , "mcsm_ppi2_outcome" , "mcsm_ppi2_affinity" , "mcsm_ppi2_scaled" , "mcsm_ppi2_outcome"
, "consurf_score" , "consurf_scaled" #, "consurf_outcome" , "consurf_score" , "consurf_scaled" , "consurf_outcome" # exists now
, "snap2_score" , "snap2_scaled" , "snap2_outcome" , "snap2_score" , "snap2_scaled" , "snap2_outcome"
, "mcsm_na_affinity" , "mcsm_na_scaled" , "mcsm_na_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{ }else{
colnames_to_extract = c(mut_colname, mut_info_colname, mut_info_label_colname colnames_to_extract = c(mut_colname, mut_info_colname, mut_info_label_colname
, aa_pos_colname, LigDist_colname , aa_pos_colname, LigDist_colname
@ -186,47 +123,29 @@ dm_om_wf_lf_data <- function(df
#======================================================================= #=======================================================================
table(comb_df_s[[mut_info_colname]]) 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 # pretty display names i.e. labels to reduce major code duplication later
foo_cnames = data.frame(colnames(comb_df_s)) foo_cnames = data.frame(colnames(comb_df_s))
names(foo_cnames) <- "old_name" names(foo_cnames) <- "old_name"
stability_suffix <- paste0(delta_symbol, delta_symbol, "G") stability_suffix <- paste0(delta_symbol, delta_symbol, "G")
flexibility_suffix <- paste0(delta_symbol, delta_symbol, "S") #flexibility_suffix <- paste0(delta_symbol, delta_symbol, "S")
lig_dn = paste0("Ligand distance (", angstroms_symbol, ")"); lig_dn #lig_dn = paste0("Ligand distance (", angstroms_symbol, ")"); lig_dn
mcsm_lig_dn = paste0("Ligand affinity (log fold change)"); mcsm_lig_dn #mcsm_lig_dn = paste0("Ligand affinity (log fold change)"); mcsm_lig_dn
lig_dn = paste0("Lig Dist(", angstroms_symbol, ")"); lig_dn
mcsm_lig_dn = paste0("mCSM-lig"); mcsm_lig_dn
duet_dn = paste0("DUET ", stability_suffix); duet_dn duet_dn = paste0("DUET ", stability_suffix); duet_dn
foldx_dn = paste0("FoldX ", stability_suffix); foldx_dn foldx_dn = paste0("FoldX ", stability_suffix); foldx_dn
deepddg_dn = paste0("Deepddg " , stability_suffix); deepddg_dn deepddg_dn = paste0("Deepddg " , stability_suffix); deepddg_dn
dynamut2_dn = paste0("Dynamut2 " , stability_suffix); dynamut2_dn dynamut2_dn = paste0("Dynamut2 " , stability_suffix); dynamut2_dn
mcsm_na_dn = paste0("mCSM-NA affinity ", stability_suffix); mcsm_na_dn mcsm_na_dn = paste0("mCSM-NA ", stability_suffix); mcsm_na_dn
mcsm_ppi2_dn = paste0("mCSM-PPI2 affinity ", stability_suffix); mcsm_ppi2_dn mcsm_ppi2_dn = paste0("mCSM-PPI2 ", stability_suffix); mcsm_ppi2_dn
consurf_dn = paste0("Consurf"); consurf_dn consurf_dn = paste0("Consurf"); consurf_dn
snap2_dn = paste0("SNAP2"); snap2_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 # change column names: plyr
new_colnames = c(asa = "ASA" new_colnames = c(asa = "ASA"
@ -235,7 +154,8 @@ new_colnames = c(asa = "ASA"
, kd_values = "KD" , kd_values = "KD"
, log10_or_mychisq = "Log10 (OR)" , log10_or_mychisq = "Log10 (OR)"
, neglog_pval_fisher = "-Log (P)" , neglog_pval_fisher = "-Log (P)"
, af = "MAF" #, af = "MAF"
, maf = "MAF"
#, ligand_dist_colname = lig_dn # cannot handle variable name 'ligand_dist_colname' #, ligand_dist_colname = lig_dn # cannot handle variable name 'ligand_dist_colname'
, affinity_scaled = mcsm_lig_dn , affinity_scaled = mcsm_lig_dn
, duet_scaled = duet_dn , duet_scaled = duet_dn
@ -245,12 +165,7 @@ new_colnames = c(asa = "ASA"
, mcsm_na_scaled = mcsm_na_dn , mcsm_na_scaled = mcsm_na_dn
, mcsm_ppi2_affinity = mcsm_ppi2_dn , mcsm_ppi2_affinity = mcsm_ppi2_dn
, consurf_score = consurf_dn , consurf_score = consurf_dn
, snap2_score = snap2_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 comb_df_sl1 = plyr::rename(comb_df_s
, replace = new_colnames , replace = new_colnames
@ -260,29 +175,26 @@ comb_df_sl1 = plyr::rename(comb_df_s
# renaming colname using variable i.e ligand_dist_colname: dplyr # 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)) comb_df_sl = comb_df_sl1 %>% dplyr::rename(!!lig_dn := all_of(ligand_dist_colname))
names(comb_df_sl) names(comb_df_sl)
#=======================
# NEW: Affinity filtered data
#========================
# mcsm-lig --> LigDist_colname
comb_df_sl_lig = comb_df_sl[comb_df_sl[[lig_dn]]<DistCutOff,]
# mcsm-ppi2 --> ppi2Dist_colname
comb_df_sl_ppi2 = comb_df_sl[comb_df_sl[[ppi2Dist_colname]]<DistCutOff,]
# mcsm-na --> naDist_colname
comb_df_sl_na = comb_df_sl[comb_df_sl[[naDist_colname]]<DistCutOff,]
##################################################################### #####################################################################
if (mut_info_label_colname == "") { static_cols1 = 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 # Selecting dfs
# with appropriate cols # with appropriate cols
#======================= #=======================
static_cols_start = c(snp_colname static_cols_start = c(snp_colname
, aa_pos_colname , aa_pos_colname
, mut_colname , mut_colname
@ -296,7 +208,8 @@ static_cols_end = c(lig_dn
, "KD" , "KD"
, "MAF" , "MAF"
, "Log10 (OR)" , "Log10 (OR)"
, "-Log (P)") #, "-Log (P)"
)
######################################################################### #########################################################################
#============== #==============
@ -312,7 +225,7 @@ expected_rows_lf = nrow(wf_duet) * (length(wf_duet) - length(pivot_cols_duet))
expected_rows_lf expected_rows_lf
# LF data: duet # LF data: duet
lf_duet = gather(wf_duet lf_duet = tidyr::gather(wf_duet
, key = param_type , key = param_type
, value = param_value , value = param_value
, all_of(duet_dn):tail(static_cols_end,1) , all_of(duet_dn):tail(static_cols_end,1)
@ -329,35 +242,6 @@ if (nrow(lf_duet) == expected_rows_lf){
wf_lf_dataL[['wf_duet']] = wf_duet wf_lf_dataL[['wf_duet']] = wf_duet
wf_lf_dataL[['lf_duet']] = lf_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 # FoldX
@ -446,7 +330,9 @@ if (nrow(lf_dynamut2) == expected_rows_lf){
# Assign them to the output list # Assign them to the output list
wf_lf_dataL[['wf_dynamut2']] = wf_dynamut2 wf_lf_dataL[['wf_dynamut2']] = wf_dynamut2
wf_lf_dataL[['lf_dynamut2']] = lf_dynamut2 wf_lf_dataL[['lf_dynamut2']] = lf_dynamut2
############################################################################
######################################################################################
#================== #==================
# Consurf: LF # Consurf: LF
#https://consurf.tau.ac.il/overview.php #https://consurf.tau.ac.il/overview.php
@ -459,9 +345,9 @@ wf_lf_dataL[['lf_dynamut2']] = lf_dynamut2
#5-->"", 6-->"", 7-->"", 8-->"", 9-->"most_conserved" #5-->"", 6-->"", 7-->"", 8-->"", 9-->"most_conserved"
#==================== #====================
# FIXME: if you add category column to consurf # FIXME: if you add category column to consurf
#cols_to_select_consurf = c(static_cols_start, c("consurf_outcome", consurf_dn), static_cols_end) 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] 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 pivot_cols_consurf = cols_to_select_consurf[1: (length(static_cols_start) + 1)]; pivot_cols_consurf
# WF data: consurf # WF data: consurf
cols_to_select_consurf = c(static_cols_start, c(consurf_dn), static_cols_end) cols_to_select_consurf = c(static_cols_start, c(consurf_dn), static_cols_end)
@ -517,15 +403,54 @@ if (nrow(lf_snap2) == expected_rows_lf){
# Assign them to the output list # Assign them to the output list
wf_lf_dataL[['wf_snap2']] = wf_snap2 wf_lf_dataL[['wf_snap2']] = wf_snap2
wf_lf_dataL[['lf_snap2']] = lf_snap2 wf_lf_dataL[['lf_snap2']] = lf_snap2
###########################################################################
# AFFINITY cols
###########################################################################
#=========================
# mCSM-lig:
# data filtered by cut off
#=========================
#---------------------
# mCSM-lig: WF and lF
#----------------------
# 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_lig[, cols_to_select_mcsm_lig] # filtered df
############################################################################ 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
#====================
# mcsm-NA affinity
# data filtered by cut off
#====================
if (tolower(gene_name)%in%geneL_na){ if (tolower(gene_name)%in%geneL_na){
#============== #---------------
# mCSM-NA: LF # mCSM-NA: WF and lF
#============== #-----------------
# WF data: mcsm-na # WF data: mcsm-na
cols_to_select_mcsm_na = c(static_cols_start, c("mcsm_na_outcome", mcsm_na_dn), static_cols_end) 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] #wf_mcsm_na = comb_df_sl[, cols_to_select_mcsm_na]
wf_mcsm_na = comb_df_sl_na[, cols_to_select_mcsm_na]
pivot_cols_mcsm_na = cols_to_select_mcsm_na[1: (length(static_cols_start) + 1)]; pivot_cols_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 = nrow(wf_mcsm_na) * (length(wf_mcsm_na) - length(pivot_cols_mcsm_na))
@ -550,14 +475,19 @@ if (tolower(gene_name)%in%geneL_na){
wf_lf_dataL[['lf_mcsm_na']] = lf_mcsm_na wf_lf_dataL[['lf_mcsm_na']] = lf_mcsm_na
} }
#-------------------------------------------------------------------
#=========================
# mcsm-ppi2 affinity
# data filtered by cut off
#========================
if (tolower(gene_name)%in%geneL_ppi2){ if (tolower(gene_name)%in%geneL_ppi2){
#============== #-----------------
# mCSM-PPI2: LF # mCSM-PPI2: WF and lF
#============== #-----------------
# WF data: mcsm-ppi2 # WF data: mcsm-ppi2
cols_to_select_mcsm_ppi2 = c(static_cols_start, c("mcsm_ppi2_outcome", mcsm_ppi2_dn), static_cols_end) 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] #wf_mcsm_ppi2 = comb_df_sl[, cols_to_select_mcsm_ppi2]
wf_mcsm_ppi2 = comb_df_sl_ppi2[, cols_to_select_mcsm_ppi2]
pivot_cols_mcsm_ppi2 = cols_to_select_mcsm_ppi2[1: (length(static_cols_start) + 1)]; pivot_cols_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 = nrow(wf_mcsm_ppi2) * (length(wf_mcsm_ppi2) - length(pivot_cols_mcsm_ppi2))
@ -582,156 +512,7 @@ if (tolower(gene_name)%in%geneL_ppi2){
wf_lf_dataL[['lf_mcsm_ppi2']] = lf_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) return(wf_lf_dataL)
} }
############################################################################ ############################################################################

4
scripts/functions/plotting_globals.R
View file

@ -39,6 +39,10 @@ resistance_col <<- "drtype"
LigDist_colname <<- "ligand_distance" LigDist_colname <<- "ligand_distance"
LigDist_cutoff <<- 10 LigDist_cutoff <<- 10
DistCutOff = 10
ppi2Dist_colname = "interface_dist"
naDist_colname = "TBC"
#================== #==================
# Angstroms symbol # Angstroms symbol
#================== #==================

2
scripts/plotting/get_plotting_dfs.R
View file

@ -112,7 +112,7 @@ cat(s1)
#source("other_plots_data.R") #source("other_plots_data.R")
#################################################################### ####################################################################
source(paste0(plot_script_path, "dm_om_data.R")) #source(paste0(plot_script_path, "dm_om_data.R"))
s2 = c("\nSuccessfully sourced other_plots_data.R") s2 = c("\nSuccessfully sourced other_plots_data.R")
cat(s2) cat(s2)

43
scripts/plotting/plotting_thesis/preformatting.R
View file

@ -10,8 +10,10 @@ source("~/git/LSHTM_analysis/config/embb.R")
source("~/git/LSHTM_analysis/scripts/plotting/get_plotting_dfs.R") source("~/git/LSHTM_analysis/scripts/plotting/get_plotting_dfs.R")
################################################################### ###################################################################
# FIXME: ADD distance to NA when SP replies # FIXME: ADD distance to NA when SP replies
# DONE: plotting_globals.R
dist_columns = c("ligand_distance", "interface_dist") dist_columns = c("ligand_distance", "interface_dist")
DistCutOff = 10 DistCutOff = 10
common_cols = c("mutationinformation" common_cols = c("mutationinformation"
, "X5uhc_position" , "X5uhc_position"
, "X5uhc_offset" , "X5uhc_offset"
@ -98,22 +100,24 @@ df3 = merged_df3
#================= #=================
# PREFORMATTING: for consistency # PREFORMATTING: for consistency
#================= #=================
df3$sensitivity = ifelse(df3$dst_mode == 1, "R", "S") # DONE: combining_dfs.R
table(df3$sensitivity) # df3$sensitivity = ifelse(df3$dst_mode == 1, "R", "S")
# table(df3$sensitivity)
# ConSurf labels # ConSurf labels
consurf_colOld = "consurf_colour_rev" #consurf_colOld = "consurf_colour_rev"
consurf_colNew = "consurf_outcome" #consurf_colNew = "consurf_outcome"
df3[[consurf_colNew]] = df3[[consurf_colOld]] #df3[[consurf_colNew]] = df3[[consurf_colOld]]
df3[[consurf_colNew]] = as.factor(df3[[consurf_colNew]]) #df3[[consurf_colNew]] = as.factor(df3[[consurf_colNew]])
df3[[consurf_colNew]] #df3[[consurf_colNew]]
# not this bit
levels(df3$consurf_outcome) = c( "nsd", 1, 2, 3, 4, 5, 6, 7, 8, 9) levels(df3$consurf_outcome) = c( "nsd", 1, 2, 3, 4, 5, 6, 7, 8, 9)
levels(df3$consurf_outcome) #levels(df3$consurf_outcome)
# SNAP2 labels # SNAP2 labels
snap2_colname = "snap2_outcome" #snap2_colname = "snap2_outcome"
df3[[snap2_colname]] <- str_replace(df3[[snap2_colname]], "effect", "Effect") #df3[[snap2_colname]] <- str_replace(df3[[snap2_colname]], "effect", "Effect")
df3[[snap2_colname]] <- str_replace(df3[[snap2_colname]], "neutral", "Neutral") #df3[[snap2_colname]] <- str_replace(df3[[snap2_colname]], "neutral", "Neutral")
# for ref: not needed perse as function already does this and assigns labels for barplots # for ref: not needed perse as function already does this and assigns labels for barplots
# labels_duet = levels(as.factor(df3$duet_outcome)) # labels_duet = levels(as.factor(df3$duet_outcome))
@ -138,14 +142,16 @@ df2 = merged_df2
#================= #=================
# PREFORMATTING: for consistency # PREFORMATTING: for consistency
#================= #=================
df2$sensitivity = ifelse(df2$dst_mode == 1, "R", "S") # DONE: combining_dfs.R
table(df2$sensitivity) # df2$sensitivity = ifelse(df2$dst_mode == 1, "R", "S")
# table(df2$sensitivity)
#---------------------------------------------------- #----------------------------------------------------
# Create dst2: fill na in dst with value of dst_mode # Create dst2: fill na in dst with value of dst_mode
# for epistasis # for epistasis
#---------------------------------------------------- #----------------------------------------------------
df2$dst2 = ifelse(is.na(df2$dst), df2$dst_mode, df2f$dst) # DONE: combining_dfs.R
# df2$dst2 = ifelse(is.na(df2$dst), df2$dst_mode, df2f$dst)
#---------------------------------------------------- #----------------------------------------------------
# reverse signs for foldx scaled values for # reverse signs for foldx scaled values for
@ -168,10 +174,11 @@ scaled_cols_stab_revised = c(scaled_cols_stab_revised, "foldx_scaled_signC")
###################################################### ######################################################
# Affinity related variables # Affinity related variables
DistCutOff = 10 # DONE:in plotting_globals.R
LigDist_colname # = "ligand_distance" # from globals # DistCutOff = 10
ppi2Dist_colname = "interface_dist" # LigDist_colname # = "ligand_distance" # from globals
naDist_colname = "TBC" # ppi2Dist_colname = "interface_dist"
# naDist_colname = "TBC"
###################################################### ######################################################
# corr colnames # corr colnames