vcf_parsing: Functions for conversion of VCF files to SQLite databases

In dbottomly/oligoMask: Tools for masking out problematic Affymetrix probes using VCF files and the oligo package

Description

Functionality to create SQLite databases from several types of VCF including those that represent the genetic diversity of populations (e.g. 1000 genomes) or inbred strains (Sanger mouse genomes project).

Usage

populate.db.tbl.schema.list(db.con, db.schema, ins.vals, use.tables = NULL, should.debug = FALSE)
make.vcf.table(db.schema, window.size, vcf.name, db.con, probe.grange, vcf.type="SNV", use.tables=NULL, limit=NULL, should.debug=FALSE, vcf.param=NULL, filter.func=NULL, filter.params=list())
create.sanger.mouse.vcf.db(vcf.files, vcf.labels, probe.tab.file, strain.names, bs.genome, db.schema, db.name="test.db", keep.category="main", window.size=1000, max.mismatch=1, limit.chr=NULL, should.debug=FALSE, package.info=NULL)

Arguments

db.con	A connection to an SQLite database.
db.schema	An object of class TableSchemaList.
ins.vals	List or other type of data to be inserted into the database using the function in the dta.func element of the TableSchemaList.
use.tables	The subset of tables the procedure should be limited to or NULL.
should.debug	Logical indicating whether additional messages should be displayed to the user
window.size	The number of aligned probes to be processed at a given time
vcf.name	Name of the VCF file
strain.names	character vector containing the genotype columns to be used
probe.grange	GRanges object containing the probe alignments information
vcf.type	Label associate with the VCF file, usually 'SNV' or 'INDEL'
limit	Number of iterations to be limited to for testing purposes
vcf.param	An object of class ScanVcfParam
filter.func	A function to be used to filter a list as returned by scanVcf For example see oligoMask:::filter.sanger.vcf. If missing or NULL, no filtering will be performed.
filter.params	A list containing named elements with any additional values to be passed to filter.func if applicable.
vcf.files	A character vector containing the path to one or more VCF files.
vcf.labels	A character vector containing the desired label for each VCF file (e.g. 'SNV' or 'INDEL').
probe.tab.file	A character vector containing the path to the tab delimited probe sequence file distributed by Affymetrix.
bs.genome	An object of class BSgenome
db.name	A character vector containing the name of the output database name.
keep.category	Categories of probes to keep from the probe.tab.file e.g. 'main'. These will be the probes that go through the SNP masking process and will be candidates for removal due to being impacted by variants.
max.mismatch	maximum number of mismatches to allow in the realignments.
limit.chr	A character vector containing the chromosomes to consider or NULL in which case all available chromosomes from the specified genome will be used.
package.info	Either NULL or a named list containing the following elements: "AUTHOR", "AUTHOREMAIL", "BOWTIE_PATH", "GENOME_PATH", "VCF_QUERY_CMD", "VCF_TYPE". If NULL is supplied, a database will be generated at the path specified in db.name. If the list is provided, a package will be generated containing the database along with other relevant metadata supplemented by the values supplied in package.info. For this list, AUTHOR and AUTHOREMAIL indicates the author of the package and their email. BOWTIE_PATH, GENOME_PATH and VCF_QUERY_CMD indicates the path to a bowtie executable, the genome used for creation of the database and the path to a vcf-query tool from vcftools or htslib, these are not necessary unless the package is to be tested afterwards and can be set to an arbitrary string. They can always be added in later. VCF_TYPE is simply a desription of the type of VCF database created for instance NODvB6 or CC.

Author(s)

Daniel Bottomly

See Also

scanVcf, TableSchemaList, ScanVcfParam, BSgenome

Examples

    if (require(BSgenome.Mmusculus.UCSC.mm9))
    {
    	vcf.files <- om.vcf.file()
    	vcf.labels <- "SNV"
    	probe.tab.file <- om.tab.file()
    	#seven strains besides the reference strain that are used in the CC
    	strain.names <- c("CASTEiJ", "AJ", "PWKPhJ", "129S1", "NZO", "NODShiLtJ", "WSBEiJ")
    	bs.genome <- BSgenome.Mmusculus.UCSC.mm9

    	#probably shouldn't do this in real analyses unless you have to
    	seqlevels(bs.genome) <- sub("chr", "", seqlevels(bs.genome))

    	db.schema <- SangerTableSchemaList()
    	db.name=tempfile()

    	keep.category <- "main"
    	window.size <- 1000
    	max.mismatch <- 0
    	should.debug <- TRUE

    	#duplications in this file, remove and recreate...
    	tab.aln <- read.delim(om.lo.file(), sep="\t", header=TRUE, stringsAsFactors=FALSE)
    	limit.chr <- GRanges(seqnames="19", ranges=IRanges(start=min(tab.aln$start), end=max(tab.aln$end)/4), strand="*")

    	create.sanger.mouse.vcf.db(vcf.files, vcf.labels, probe.tab.file, strain.names, bs.genome, db.schema, db.name, keep.category, window.size, max.mismatch, limit.chr, should.debug)

    	db.con <- dbConnect(SQLite(), db.name)

    	var.ovls <- dbGetQuery(db.con, "SELECT * FROM probe_to_snp NATURAL JOIN reference NATURAL JOIN probe_align NATURAL JOIN probe_info limit 5")

    	var.ovls
      }

dbottomly/oligoMask documentation built on May 15, 2019, 1:22 a.m.