======================
A guide to using khmer
======================

khmer can be used for several different assembly-related functions at this
point.  What you need or use will depend on what kind of data you have.

Note that khmer was built around Illumina sequences: short reads under ~150bp
in length.  Longer sequences will screw up various assumptions; check with
the authors to see how, or read the source code :).

Metagenomes
===========

You'll probably be interested in:

 - low-abundance k-mer trimming; see filter-abund.py in :doc:`scripts`.
 - partitioning
 - digital normalization; see normalize-by-median.py in :doc:`scripts`.

mRNAseq
=======

 - low-abundance k-mer trimming; see filter-abund.py in :doc:`scripts`.
 - digital normalization; see normalize-by-median.py in :doc:`scripts`.

Genomes
=======

khmer's graph-filtering functionality can be used to remove repeats from
raw reads, which for at least some data sets scales the assembly process
rather dramatically.

ChIP-seq
========

We've never used khmer for ChIP-seq analysis, sorry!