In DataConversion.R, we could add this code which would covert bed to fastSK fasta format. It relies heavily on genNullSeqs from gkmSVM so we would need to include downloads for gkmSVM.
"""
library(gkmSVM)
#' @name convertFromBed
#' @description Converts a bed file into positive and negative sequence format used to train fast-deltaSVM. Converts a bed file into positive and negative sequence format used to train fast-deltaSVM.
#' Most parameters are from genNullSeqs which is part of gkmSVM, while the train_out_file is the input to fast-deltaSVM.
#' @export
convertFromBed <- function(inputBedFN, genomeVersion, outputBedFN, outputPosFastaFN, outputNegFastaFN, train_out_file, xfold=1, repeat_match_tol=0.02, GC_match_tol=0.02, length_match_tol=0.02, batchsize=5000, nMaxTrials=20, genome=NULL){
genNullSeqs(inputBedFN, genomeVersion, outputBedFN, outputPosFastaFN, outputNegFastaFN, xfold, repeat_match_tol, GC_match_tol, length_match_tol, batchsize, nMaxTrials, genome);
convertFromGKM(outputPosFastaFN, outputNegFastaFN, train_out_file);
}
"""
In DataConversion.R, we could add this code which would covert bed to fastSK fasta format. It relies heavily on genNullSeqs from gkmSVM so we would need to include downloads for gkmSVM.
"""
library(gkmSVM)
#' @name convertFromBed
#' @description Converts a bed file into positive and negative sequence format used to train fast-deltaSVM. Converts a bed file into positive and negative sequence format used to train fast-deltaSVM.
#' Most parameters are from genNullSeqs which is part of gkmSVM, while the train_out_file is the input to fast-deltaSVM.
#' @export
convertFromBed <- function(inputBedFN, genomeVersion, outputBedFN, outputPosFastaFN, outputNegFastaFN, train_out_file, xfold=1, repeat_match_tol=0.02, GC_match_tol=0.02, length_match_tol=0.02, batchsize=5000, nMaxTrials=20, genome=NULL){
genNullSeqs(inputBedFN, genomeVersion, outputBedFN, outputPosFastaFN, outputNegFastaFN, xfold, repeat_match_tol, GC_match_tol, length_match_tol, batchsize, nMaxTrials, genome);
convertFromGKM(outputPosFastaFN, outputNegFastaFN, train_out_file);
}
"""