VcfBasicRules-class: VCF filters class objects sub-types
In TVTB: TVTB: The VCF Tool Box

Description Details Accessor methods Constructors Subsetting and Replacement Combining Evaluating Author(s) See Also Examples

The VcfFixedRules and VcfInfoRules classes store filters applicable to the fixed and info slots of VCF objects, respectively.

The VcfVepRules stores filters applicable to Ensembl VEP predictions stores in a given INFO key.

All arguments are first passed to S4Vectors::FilterRules before re-typing the resulting as a VcfFixedRules, VcfInfoRules, or VcfVepRules class.

In the following code snippets x is an object from any of the classes decribed in this help page, except when specified otherwise.

active(x), active(x)<-: Gets or sets the active state of each filter rule in x. Inherited from FilterRules
vep(x), vep(x)<-: Gets or sets the INFO key where the Ensembl VEP predictions to use for filtering are stored. Returns NA_character_ for filters not applicable to VEP predictions.
type(x): Returns "filter" (linkS4class{FilterRules}), "fixed" (linkS4class{VcfFixedRules}), "info" (linkS4class{VcfInfoRules}), or "vep" (linkS4class{VcfVepRules}) as a character vector of length(x).

VcfFixedRules(exprs = list(), ..., active = TRUE)

VcfInfoRules(exprs = list(), ..., active = TRUE)

VcfVepRules(exprs = list(), ..., active = TRUE, vep = "CSQ")

All methods construct an object of the corresponding class with the rules given in the list exprs or in .... The initial active state of the rules is given by active, which is recycled as necessary.

See the constructor of FilterRules for more details.

In the following code snippets x and value are objects from any of the classes described in this help page.

: x[i]: Subsets the filter rules using the same interface as for List.
: x[[i]]: Extracts an expression or function via the same interface as for List.
: x[i] <- value: Replaces a filter rule by one of the same class. The active state(s) and name(s) are transferred from value to x.
: x[[i]] <- value: The same interface as for List. The default active state for new rules is TRUE.

In the following code snippets x, values, and ... are objects from any of the classes described in this help page, or VcfFilterRules.

: append(x, values, after = length(x)): Appends the values onto x at the index given by after.
: c(x, ...,): Concatenates the filters objects in ... onto the end of x.

Note that combining rules of different types (e.g. VcfFixedRules and VcfVepRules) produces a VcfFilterRules object.

As described in the S4Vectors documentation:

eval(expr, envir, enclos): Evaluates a rule instance (passed as the expr argument) in their respective context of a VCF object (passed as the envir argument). i.e.:

VcfFixedRules: fixed(envir)
VcfInfoRules: info(envir)
VcfVepRules: mcols(parseCSQToGRanges(envir, ...))
FilterRules: envir

evalSeparately(expr, envir, enclos):

subsetByFilter(x, filter)

summary(object)

See eval,FilterRules,ANY-method for details.

Kevin Rue-Albrecht

FilterRules, VcfFilterRules, and VCF.

# Constructors ----

fixedRules <- VcfFixedRules(list(
    pass = expression(FILTER == "PASS"),
    qual = expression(QUAL > 20)
    ))
fixedRules

infoRules <- VcfInfoRules(list(
    common = expression(MAF > 0.01), # minor allele frequency
    alt = expression(ALT > 0) # count of alternative homozygotes
    ))
infoRules

vepRules <- VcfVepRules(list(
    missense = expression(Consequence %in% c("missense_variant")),
    CADD = expression(CADD_PHRED > 15)
    ))
vepRules

filterRules <- S4Vectors::FilterRules(list(
    PASS = function(x) fixed(x)$FILTER == "PASS",
    COMMON = function(x) info(x)$MAF > 0.05
    ))
filterRules

# Accessors ----

## get/set the active state directly
S4Vectors::active(infoRules)
S4Vectors::active(infoRules)["common"] <- FALSE

## See S4Vectors::FilterRules for more examples


# Example data ----

# VCF file
vcfFile <- system.file("extdata", "moderate.vcf", package = "TVTB")

# TVTB parameters
tparam <- TVTBparam(Genotypes("0|0", c("0|1", "1|0"), "1|1"))

# Pre-process variants
vcf <- VariantAnnotation::readVcf(vcfFile, param = tparam)
vcf <- VariantAnnotation::expand(vcf, row.names = TRUE)
vcf <- addOverallFrequencies(vcf)


# Applying filters to VCF objects ----

## Evaluate filters
S4Vectors::eval(fixedRules, vcf)
S4Vectors::eval(infoRules, vcf)
S4Vectors::eval(vepRules, vcf)
S4Vectors::eval(filterRules, vcf)

summary(S4Vectors::eval(vepRules, vcf))


## Evaluate filters separately
S4Vectors::evalSeparately(vepRules, vcf)

summary(S4Vectors::evalSeparately(vepRules, vcf))

## Subset VCF by filters
S4Vectors::subsetByFilter(vcf, vepRules)


# Subsetting and Replacement ----

vep1 <- vepRules[1] # VcfVepRules
vepRules[[1]] # expression

# Replace the expression (active reset to TRUE, original name retained)
vepRules[[2]] <- expression(CADD_PHRED > 30)

# Replace the rule (active state and name transferred from v5obj)
vepRules[2] <- VcfVepRules(
    list(newRule = expression(CADD_PHRED > 30)),
    active = FALSE)