priorProcess: Process DNase-seq (and histone ChIP-seq) datasets

In yzheng74/permseq: Mapping protein-DNA interactions in highly repetitive regions of the genomes with prior-enhanced read mapping

Description

Process DNase-seq and/or histone ChIP-seq data and construct required DNase-seq and histone ChIP-seq files for generating priors to allocate multi-reads in ChIP-Seq data.

Usage

priorProcess(dnaseFile = NULL, histoneFile = NULL, dnaseName
= "dnase", histoneName = NULL, fragL = 200, AllocThres
= 900, chrList = NULL, capping = 0, outfileLoc = "./",
outfile = "dnase_histone", bowtieDir, bowtieIndex, vBowtie = 2,
mBowtie = 99, pBowtie = 8, bwaDir, bwaIndex, nBWA = 2, oBWA = 1, tBWA =
8, mBWA = 99, csemDir, picardDir, chrom.ref, saveFiles = TRUE)

Arguments

dnaseFile	DNase-seq file in fastq format. For faster results, sam formatted file after alignment including multi-mapping reads or bam or bed files already obtained by CSEM with allocated reads can also be processed, if available. Otherwise, it is better start from the fastq formatted file. The default value is NULL.
histoneFile	Histone ChIP-seq file in fastq format. For faster results, sam formatted file after alignment including multi-mapping reads or bam or bed files already obtained by CSEM with allocated reads can also be processed, if available. Otherwise, it is better start from the fastq formatted file. Default value is set to NULL.
dnaseName	Name of DNase-seq data or the dataset used as DNase-seq data in the model, default set as 'dnase'.
histoneName	Name of histone ChIP-seq data sets. If no giving values, histoneName is set as a vector of index number 1:length(histoneFile).
fragL	Average fragment length with default value 200.
AllocThres	Allocation threshold. It will select reads with scores higher than AllocThres (allocation probability*1000). Default set at 900.
chrList	A vector of chromosomes that will be included in the analysis. Default set as NULL and priorProcess will get the list from processed files. Otherwise, if given by the user, it should be consistent with the chromosome name(s) in the corresponding fasta file(s). For more information, see details.
capping	Maximum number of reads allowed at each nucleotide position. To avoid potential PCR amplification artifacts, the maximum number of reads that can start at a nucleotide position is capped. Default is 0 (no capping, i.e. no maximum restriction).
outfileLoc	Directory to store processed files.
outfile	Infix of outfile name. Default set as "dnase_histone" indicating the prior is constructed using DNase-seq and Histone data.
bowtieDir	Directory where Bowtie was installed, default set as NULL.
bowtieIndex	Bowtie index, used in bowtie aligning. Default value is NULL and users can specify the selection of aligner, Bowtie or BWA, by providing the corresponding index.
vBowtie	Bowtie parameter. In -v mode, alignments may have no more than vBowtie mismatches, where v may be a number from 0 through 3 set using the -v option. Default value is 2.
mBowtie	Bowtie parameter. -m parameter instructs bowtie to refrain from reporting any alignments for reads having more than mBowtie reportable alignments. Default value is 99 allowing multi-reads alignment.
pBowtie	Bowtie parameter. The -p option causes Bowtie to launch a specified number of parallel search threads. Each thread runs on a different processor/core and all threads find alignments in paralle. Default value is 8.
bwaDir	Directory where BWA was installed. Default set as NULL.
bwaIndex	BWA index used in BWA alignment. Users can specify the aligner, Bowtie or BWA, by specifying the index that will be used. Default set as NULL.
nBWA	BWA paramter. In "bwa aln -n" mode, if it is an integer, it denotes the maximum edit distances including mismatch and gap open. Otherwise, it will be the fraction of missing alignments given 2% uniform base errr rate. Default value is 2.
oBWA	BWA parameter. In "bwa aln -o" mode, it specifies the maximum number of gap open. Default set as 1.
tBWA	BWA parameter. In "bwa aln -t" mode, it is the number of threads in multi-threading mode. Default set as 8.
mBWA	BWA parameter. In "bwa samse -n", it restricts the maximum number of alignments to output for each read. If a read has more hits, the XA tag will not be written. Default set as 99.
csemDir	Directory where CSEM was installed.
picardDir	Directory where PICARD jar file is saved.
chrom.ref	reference genome index summary information. First line is the number of chromosomes in the index, either bwaIndex or bowtieIndex, including chrM. Second line is the size of each chromosome. Third line is the name of chromsome.
saveFiles	Option to save intermediate files created. Default set as TRUE.

Details

Processes DNase-seq and/or histone ChIP-seq files and generates module for further analysis in priorGenerate. If no DNase-seq data available and do not know which histone data could play as the DNase-seq data in the model, start from priorHistone_init and priorHistone_multi functions.

If no DNase-seq or histone ChIP-seq data available, run readAllocate directly and multi-reads will be allocated without using prior information.

If no chrList is provided, priorProcess will generate the list from processed files (.sam file if DNase-seq input file is in fastq format or .bed file if DNase-seq input file is in .bam or .bed format). Otherwise, if given by the user, it will accelerate the procedure, but the chrList should be consistent with the chromosome name(s) in the corresponding .fa or .fasta file(s). In other words, for example, it should be the name on the first line after ">" in .fa file.

Users can select from Bowtie and BWA to do the alignment by providing the corresponding index and leaving the other as default value NULL. If both indices are provided, the package will automatically use Bowtie to do the multi-mapping reads alignment.

DNase-seq and/or Histone aligned sam file will go through filtering process to remove duplicates. By default, 'samtools rmdup -s' function will be used. PICARD jar can take over if PICARD jar path is provided.

plot(), summary(), names() and print() methods can be used to see the information contained in "Prior" object. To obtain the ChIP-drq alignment information from bowtie, use summary().

Value

A new "Prior" object is created containing the following information:

dnaseName	Name of dataset that is used as DNase-seq, especially in the "Only histone" situation, dnaseName is the selected histone ChIP-seq dataset.
dnaseAlign	DNase-seq alignment summary information from bowtie.
dnaseKnots	A vector of knots points for the B-spline functions. They are the 90, 99 and 99.9th percentiles of read counts.
dnaseThres	A vector of DNase-seq group created to generate aggregated ChIP data. After alignment, positions which have the same DNase-seq read count are clustered into one group. dnaseThres is the corresponding read count number in each group. Each count is corresponding to one group and the grouping data partitions the whole genome into multiple segments.
posLoc_bychr	Location of the files containing the group index of each segment of the genome.
histoneName	Name of histone ChIP-seq dataset(s). If no giving values, the histoneName would be set as a vector of index number(1:length(histoneFile)).
histoneNum	Number of histone ChIP-seq dataset(s).
histoneAlign	histone ChIP-seq alignment summary information from bowtie.
dataNum	Number of dataset(s) that are used.
chrList	Chromosome list.
fragL	Fragment length.
bowtieInfo	List of Bowtie related information: bowtieDir, bowtieIndex, vBowtie, mBowtie and pBowtie.
bwaInfo	List of BWA related information: bwaDir, bwaIndex, nBWA, oBWA, tBWA, mBWA.
csemDir	Directory where CSEM was installed.
picardDir	Directory where PICARD jar file is saved.
outfileLoc	Directory to store output files.
chrom.ref	Name of the file for chromosome info.

Author(s)

Xin Zeng, M. Constanza Rojo-Alfaro, Ye Zheng

Examples

## Not run: 
object = priorProcess(dnaseFile = NULL, histoneFile = NULL, dnaseName = 'dnase',
histoneName = NULL, fragL = 200, chrList, capping = 0, outfileLoc = "./",
outfile, bowtieIndex, csemDir, picardDir, chrom.ref, saveFiles = FALSE)

## End(Not run)

yzheng74/permseq documentation built on May 4, 2019, 8:47 p.m.