Standalone genefeatures program.

This program scans a sequence in FASTA format for localised gene 
features.  These features are then written to stdout in GFF format.

Usage: genefeatures [opts] <tablefile> <sequencefile>
eg     genefeatures tables b0334.dna

  -h(elp)      : print this message
  -v(ersion)   : print genefeatures version info
  -segs <n>    : cutoff for segment scores (default 1.0)
  -splice3 <n> : cutoff for splice3 scores (default -2.0)
  -splice5 <n> : cutoff for splice5 scores (default 0.0)
  -atg <n>     : cutoff for ATG scores (default 0.0)
  -stop <n>    : cutoff for stop scores (default -2.0)

To predict features, you give it sequence profiles of the features 
that you're interested in.  Examples of these are included with 
this distribution, eg humtab.atg. 

The <tablefile> parameter is the name of a file containing a list 
of all the profile files to be used.

In fact, you make two files for each feature: one representing a profile of
positive examples of the feature of interest (e.g. real splice sites), and
the other a profile of identical dimensions made from "pseudo" examples of
the feature (e.g. pieces of sequence that have some splice-site
characteristics, but are not splice-sites.  This second table is most often
made by just making a profile from random pieces of DNA, hence the
presence of table names that start with "ranseq").

Although the gene_features binary can be anywhere you like, you must be 
in the directory where the profile and sequence files exist when you 
run the program, or it won't be able to find the data to process.

The <sequencefile> holds the FASTA-format sequence you want to scan, 
of course.

To build a new binary from the source.

A Makefile is provided, so as long as you have all the source code,
you should just be able to type "make".

The files included with this distribution are as follows:

Makefile
README
gfcode.c
readseq.c
readseq.h
regular.h
genefeatures