unparalog <- function(DATA, paralog_separator = ";", annotation_separator = ",", GeneColName , AnnotationColName ){
# Installing missing dependencies
dependencies <- c("stringi", "progress")
missing_packages <- dependencies[!(dependencies %in% installed.packages()[, "Package"])]
if(length(missing_packages)) install.packages(missing_packages)
# Sanity checks
check_paralog_sep <- !any(stringi::stri_detect_fixed(str = DATA$Gene.refGene,pattern = paralog_separator))
check_annotation_sep <- !any(stringi::stri_detect_fixed(str =  DATA$AAChange.refGene, pattern = annotation_separator))
idx_Gene <- which(colnames(DATA) == GeneColName)
idx_Annotation <- which(colnames(DATA) == AnnotationColName)
if(length(idx_Gene) == 0 | length(idx_Annotation) == 0 | check_paralog_sep | check_annotation_sep) {
message(stringi::stri_c("You entered :-->",
"\nparalog_separator = ",paralog_separator,
"\nannotation_separator = ",annotation_separator,
"\nGeneColName = ", GeneColName,
"\nAnnotationColName = ",AnnotationColName))
stop("Inconsistencies with these input parameters.\n Ensure they are correct and try again.")
}
# Cleanup
rm(missing_packages,dependencies,check_paralog_sep,check_annotation_sep)
gc()
current.idx <- nrow(DATA)+1
paralog.idx <- which(stringi::stri_detect_fixed(str = DATA$Gene.refGene,pattern = paralog_separator))
pb <- progress::progress_bar$new(total=length(paralog.idx),format = " [:bar] :current/:total (:percent)",); pb$tick(0)
Number_of_paralogs <- sum(stringi::stri_count_fixed(str = DATA$Gene.refGene,pattern = paralog_separator))
message(stringi::stri_c("There are ",length(paralog.idx)," annotations with ", Number_of_paralogs," paralogs."))
# copying structure of original DATA
DATA.add <- DATA[1,]; DATA.add <- DATA.add[-1,]
# Adding the empty rows to original table.
# These rows will be populated in the for loop
DATA.add <- dplyr::bind_rows(DATA.add,data.frame(matrix(nrow = (length(paralog.idx)+Number_of_paralogs), ncol = ncol(DATA),dimnames = list(c(),colnames(DATA))),stringsAsFactors = F))
DATA <- dplyr::bind_rows(DATA,DATA.add)
rm(DATA.add) ; gc()
# Beginning isolation of the paralogs
for(i in paralog.idx){
Muts <- unlist(stringi::stri_split_fixed(DATA$AAChange.refGene[i],annotation_separator),use.names = F,recursive = F)
for (gene in unlist(stringi::stri_split_fixed(DATA$Gene.refGene[i],pattern = paralog_separator),use.names = F,recursive = F)){
DATA[current.idx,] <- DATA[i,]
DATA$Gene.refGene[current.idx] <- gene
DATA$AAChange.refGene[current.idx] <- paste0(Muts[grep(gene,Muts,fixed = T)],collapse=",")
current.idx <- current.idx+1
}
pb$tick(1)
}
# removing the original rows with the paralogs as they are all unparalogged now
DATA <- DATA[-paralog.idx,]
rownames(DATA) <- as.character(seq(1,length(DATA[,1])))
return (DATA)
}
dependencies <- c("stringi")
missing_packages <- dependencies[!(dependencies %in% installed.packages()[, "Package"])]
if(length(missing_packages)) install.packages(missing_packages)
rm(missing_packages,dependencies)
MutSiteFind <- function(MutationColumn){
# #<---------------------------->
# # Please include this section when distributing and/or using this code.
# # Please read and abide by the terms of the included LICENSE.
# # Copyright 2020, Deepankar Chakroborty
# #
# #  Author : Deepankar Chakroborty (https://gitlab.utu.fi/deecha)
# #  Report issues: https://gitlab.utu.fi/deecha/shared_scripts/-/issues
# #  License: https://gitlab.utu.fi/deecha/shared_scripts/-/blob/master/LICENSE
# #
# #  PURPOSE:
# #  For a given vector of amino acid changes like A123T, V256F, E746_A750del
# #  this function returns c(123, 256, 746) as amino acid positions of
# #  the mutated residue.
# #     In case of indels, it doesn't return the range!! (returns only the start position)
# #
# #<---------------------------->
return(unlist(x = stringi::stri_extract_first_regex(str = MutationColumn,pattern = "[[:digit:]]+"), use.names = F))
}
MutSiteFind