KOBAS - KEGG Orthology Based Annotation System

1. What's KOBAS?

KEGG Orthology Based Annotation System (KOBAS) is a standalone Python
application in Bioinformatics.  KOBAS can assign appropriate KO terms for
queried sequences based on similarity search, and it can further discover
enriched KO terms among the annotation results by frequency of pathways or
statistical significance of pathways.

2. Installation

Please refer to install.txt under docs directory in KOBAS source

3. Usage

KOBAS composes of two major scripts, which are similar to usual command line
tools: the behavor can be adjusted by swiches and configuration file.  The whole
work flow can be divided to two steps: sequence annotation and pathway discovery.

Sequence Annotation
-----------------

Given a FASTA file of protein, you can annotate them by this command:

      blast2ko.py <FASTA FILE> 

All the result are printed default, you can redirect into a file.  If the input
are nuclear acid sequences, you can adjust blast2ko behavor by the command
option "-p" as follows:

      blast2ko.py -p blastx <FASTA FILE>


Pathway Discovery
-----------------

Once the annotation is complete, pathfind can be used to discover enriched
pathways by two ways: frequency prior and statistally significance prior.
You can find the most frequent pathways among the annotation result:

      pathfind.py <ANNOATION FILE>

Or, you can also find the most statistically enriched pathways:

      pathfind.py [-k b/c/h] <ANNOATION FILE> <BACKGROUND ORGANISM>

'-k' option specifies the statistical method to rank, 'b' represents binomial
distribution, 'c' represents chisquare distribution and 'h' represents
hypergeometry distrubtion. <BACKGROUND Organism> is three-charater abbreviation
of the organism in KEGG GENES as backgroud distribution.  Also, you can specifiy
an data file as backround like this:

      pathfind.py [-k b/c/h] <ANNOATION FILE> <BACKGROUND FILE>

Please refer other details to online manpage by the command "<CMD> -h".

4. Contact

Please, report bugs, comments and suggestions to the email below.  Also, if you
succeed in compiling this module under a platform other than Linux, please, send
us a note.

Authors: Xizeng Mao, Jianmin Wu, Tao Cai, Chen Xie, Liping Wei

Email:  kobas@mail.cbi.pku.edu.cn

Web: 	http://kobas.cbi.pku.edu.cn

Enjoy :)