Conditionals and Logic

Included in PopGen.jl are some functions to help discriminate your data a bit more. Like all conditionalsℹ️, these return true or false depending on the test.

ℹ️ Missing values

By Julia's design, conditionals on missing values return missing. For indexing and subsetting reasons, ishom and ishet return false on missing values, however unexported methods _ishom and _ishet return missing as per the standard convention. These unexported methods are critical for calculations where missing values should absolutely not be treated as false.

Homozygosity

ishom(locus::Genotype)
ishom(locus::GenoArray)

This will return true if a genotype is homozygous. The GenoaArray version just broadcasts it across all the genotypes in an array, returning a vector of true or false.

Example

julia> cats = @nancycats ;
julia> subset = cats[1:10, :genotype]
10-element Vector{Union{Missing, Tuple{Int16, Int16}}}:
 missing
 missing
 (135, 143)
 (133, 135)
 (133, 135)
 (135, 143)
 (135, 135)
 (135, 143)
 (137, 143)
 (135, 135)

julia> ishom(subset[3])
false

julia> ishom(subset)
10-element Vector{Bool}:
 0
 0
 0
 0
 0
 0
 1
 0
 0
 1

using skipmissing

If you want to avoid missing genotypes, you can use skipmissing to ignore them. This also works for ishet.

julia> ishom(skipmissing(subset))
8-element Vector{Bool}:
 0
 0
 0
 0
 1
 0
 0
 1

Another option is to check if a genotype is homozygous for a specific allele. To do that, we exploit Julia's multiple dispatch and use ishom again, but with different arguments.

ishom(geno::Genotype, allele::Signed)
ishom(genos::GenoArray, allele::Signed)

This will return true if the geno (or genos) is/are homozygous for the specified allele. Notice that when we query a genotype that doesn't contain that allele, it returns false.

Example

julia> ishom(subset[3], 135)
false

julia> ishom(subset[10], 135)
true

julia> ishom(subset[9], 135)
false

julia> ishom(subset, 135)
10-element Vector{Bool}:
 0
 0
 0
 0
 0
 0
 1
 0
 0
 1

Heterozygosity

ishet(locus::Genotype)
ishet(locus::GenoArray)

This is the exact opposite of ishom, returning true if the genotype (or genotypes) is/are heterozygous.

Example

julia> cats = @nancycats ;
julia> subset = cats[1:10, :genotype]
10-element Vector{Union{Missing, Tuple{Int16, Int16}}}:
 missing
 missing
 (135, 143)
 (133, 135)
 (133, 135)
 (135, 143)
 (135, 135)
 (135, 143)
 (137, 143)
 (135, 135)

julia> ishet(subset[3])
true

julia> ishet(subset)
10-element Vector{Bool}:
 0
 0
 1
 1
 1
 1
 0
 1
 1
 0

We likewise have the option to check if a locus is heterozygous for a specific allele. To do that, we again exploit Julia's multiple dispatch and use ishet, but with different arguments.

ishet(geno::Genotype, allele::Signed)
ishet(genos::GenoArray, allele::Signed)

This will return true if the geno (or genos) is/are heterozygous for the specified allele. Notice that when we query a genotype that doesn't contain that allele, it returns false.

Example

julia> ishet(subset[3], 135)
true

julia> ishet(subset[10], 135)
false

julia> ishet(subset[9], 135)
true

julia> ishet(subset, 135)
10-element Vector{Bool}:
 0
 0
 1
 1
 1
 1
 0
 1
 0
 0

Biallelic data

Some analyses are restricted to work exclusively on biallelic data (e.g. Hudson pairwise FST), so it may help to know if things are biallelic.

isbiallelic(data::GenoArray)

Returns true if the GenoArray is biallelic, false if not.

isbiallelic(data::PopData)

Returns true all the loci in the PopData are biallelic, false if not.

Example

julia> sharks = @gulfsharks ;

julia> isbiallelic(sharks)
false

julia> drop_multiallelic!(sharks)
[ Info: Removing 258 multialleic loci

julia> isbiallelic(sharks)
true