NullHap - quick start

The NullHap utility estimates haplotype frequencies using genotype data from unrelated individuals. It can be applied to loci with a codominant allele(s) and/or a silent (null) alleles (as often happens with KIR loci). The NullHap is based on maximum likelihood approach and uses an expanded Expectation-Maximalisation algorithm.


NullHap [options] input_file

Allowed options:

example input file

Details of input file format are given below. Two sections should be given : description of loci and an description of observations. The ';' (semicolon) starts a comment. It ends at the next newline.

Loci description section contains description of every locus. The keyword 'Locus' (or 'L') starts an locus description. Next, in parentheses three locus attributes should be defined:

that is: Locus("name" variants_number has_null) Example:
    Locus("A" 2 true)
    Locus("B" 2 false)

Observation descriptions section contains the genotypes and number of observations. The keyword 'Observation' (or 'O') starts an observation description, then two observation attributes should be defined:

The genotype is a sequence of observed alleles separated by a dot ('.'). The allele name is the locus name followed by variant number. The null allele is described by locus_name0 (for example A0, if the locus name is A)

Null allele should be included only if no other allele was found in a given locus (i.e. in a homozygote). So: Observation("genotype" number_of_individuals) Examples:

    O("A1.B1." 3)
    O("A1.B1.B2." 5)
    O("A0.B1." 1)

The results are send to the standard output. It is a set of lines, each containing a haplotype description and its probability. When the results should be written to a file type:

    NullHap input_file > output_file

main page