Generates a random sibling from a given DNA profile from a given set of
allele frequencies. At each locus, the sibling inherits the first allele of
the given profile with one quarter, or the second allele with probability
one quarter, both alleles with probability one quarter, or neither with
probability one quarter. If the sibling inherits zero or one identical
alleles, the missing alleles are chosen at random with probability
proportional to the allele frequencies.