mlgt (version 0.16)

callGenotypes: Make genotype calls

Description

Apply a genotype call method to a table or list of tables of variant data such as the markerSampleList table of an mlgtResult.

Usage

callGenotypes(resultObject, method = "callGenotypes.default", markerList = names(resultObject@markers), sampleList = resultObject@samples, mapAlleles = FALSE, alleleDb = NULL, approxMatching = FALSE, ...)

Arguments

resultObject

An object of class mlgtResult, as returned by mlgt

alleleDb

A list of variantMap objects derived from known alleles. As made by createKnownAlleleList

method

How to call genotypes. Currently only "callGenotypes.default" is implemented. Users can define their own methods as R functions (see the vignette).

markerList

For which of the markers do you want to call genotypes (default is all)?

sampleList

For which of the samples do you want to call genotypes (default is all)?

mapAlleles

FALSE/TRUE. Whether to map variants to db alleleDb of known alleles.

approxMatching

If TRUE, a BLAST search is also performed to find matches (slower). Additional columns are added to the genoytpeTable

...

Other parameter values will be passed to custom methods such as callGenotypes.default

Value

list of call results including the call parameters and a table of calls (class genotypeCall). If an mlgtResult object was supplied then a list of genotypeCall objects will be returned, each named by marker.

Details

After mlgt has generated tables of the most common variants assigned in each marker/sample pair, an attempt can be made to call genotypes. This is kept separate because users might want to try different calling methods and have the option to map to a known set of alleles. Currently, only one method is implemented (`custom'). See callGenotypes.default. This function also includes the option to map variants to a list of known alleles created using createKnownAlleleList. The basic method makes only perfect matches but a secondary method can be triggered (approxMatching=TRUE) to find the allele with the greatest similarity using a local BLAST search.

Examples

## Not run: 
# data("mlgtResult", package="mlgt")
# my.mlgt.Result
# # the default method
# my.genoytpes <- callGenotypes(my.mlgt.Result)
# # using a custom method
# callGenotypes.custom <- function(table, maxPropUniqueVars=0.5) {
# 	table$status <- "notCalled"
# 	table$propUniqueVars <- table$numbVar/table$numbSeq
# 	table$status <- ifelse(table$propUniqueVars <= maxPropUniqueVars,"good", "bad")
# 	return(table)
# }
# my.custom.Genotypes <- callGenotypes(my.mlgt.Result, method="callGenotypes.custom")
# ## End(Not run)