It's hard to install original version of dnase2tf in windows system, so i put dnase2tf on github to install it handily by:
devtools::install_github('jiang-junyao/dnase2tf')
For more detail of this package, please see: https://sourceforge.net/projects/dnase2tfr/
dnase2tf(datafilepath, hotspotfilename, mapfiledir, outputfilepath, ...) finds footprint candidates from the DNase-I seq data (given by datafilepath) on the user specified regions of interest (given by hotspotfilename). Mappability file location (mapfiledir) and the filename for the output file are also required.
Datafilepath: the directory path+ the file name prefix of the DNaseI sequence reads. The input files are split corresponding to each chromosome and shares the common "prefixes" and 'suffixes' in their file names. Those names differ between chromosomes. These files are assumed to be in the format 'DATA NAME(Prefix)' + '_chr#' + '.txt'. For example, typical data files would be named as "DHS_Dex_chr1.txt?, 'DHS_Dex_chr2.txt', etc and located in the same folder. In the data files, each row represents the genomic location of a mapped DNA sequence (chromosome, start, stop, strand). Tabs separate fields and coordinates are 1-based.
Hotspotfilename: the directory path of the region of interest in BED file format (0-based).
Mapfiledir: the directory path of the mappability file of the reference genome. The program assumes that mappability files named by chromosomes as 'chr1b.txt', 'chr2b.txt', etc.
Outputfilepath: the directory path + the file name prefix of the output. This program outputs in both BED and BEDGraph file formats with various FDR (False Discovery Ratio) starting from 0% to 100%.
Options
assemseqdir : the directory path of the assembly sequence files (chr1.fa, chr2.fa, etc..) which can be downloaded from the following links: http://hgdownload.soe.ucsc.edu/goldenPath/hg19/bigZips/chromFa.tar.gz (hg19) http://hgdownload.soe.ucsc.edu/goldenPath/mm9/bigZips/chromFa.tar.gz (mm9) If omitted, dinucleotide bias correction is disabled in the computation.
dftfilename: the dinucleotide bias percentage (DFT) file. A DFT file is a tab-deliminated text file containing expected and observed proportions dinucleotide DNAs measure from the data. The source code of the DFT file generation program is also provided with this package and the user may compile and run the program. This option is ignored if dinucleotide bias correction is not used.
numworker: the number of CPU cores to run concurrently. To prevent the problem caused by out of memory error, try the small number first. By default, 2 workers are assigned for the computation.
maxw: The maximum width of footprint candidates. '30' by default.
minw: The minimum width of footprint candidates. '6' by default.
z_threshold: z-score threshold of initial footprint candidate selection. '-2' is set to a default value.