hapmotif is a program for detecting conserved sequence motifs in a set
of haploid genetic sequences.  It is based on a combination of dynamic
programming and expectation maximization algorithms for parsing
sequences in terms of conserved motifs and refining motif frequencies
based on the parses.  The algorithms are described in the paper
"Haplotype Motifs: An Algorithmic Approach to Locating Evolutionarily
Conserved Patterns in Haploid Sequences" by Russell Schwartz, which
will appear in the proceedings of the 2003 IEEE Computer Society
Conference on Computational Biology.

The program is called as follows:

hapmotif [-i <input file> ] [-o <output format>]
         [-m <mutation penalty>] [-c <confidence interval>]
         [-x <maximum motif length>]

<input file> is the reference file name (default: standard input)
<output format> is one of txt, html, and pnm (default: txt)
<mutation penalty> is a prior probability of mutation (default: 0)
<confidence interval> is the size of a confidence region used to
establish which sequences are statistically over-represented
(default: 0.999)
<maximum motif length> is the longest motif for which the program
will search (default 20)

The program takes as input a file describing the variant sites in a
set of haploid sequences.  It produces one of three forms of output
describing these sequences in terms of conserved motifs made up of
alleles at consecutive variable sites.

See README.files for a description of the input file format and the
output file formats and options.  See the paper for a description of
the methodology and the precise meaning of the parameters.

Note that the statistical model has been changed from the original
implementation.  During initial motif discovery, a motif's frequency
is estimated to be the excess of its measured frequency beyond the
frequency of its best explanation in terms of smaller motifs.
Previously, a motif's frequency was initially estimated to be its
measured frequency.