Generates DNA sequences using a Markov Chain, learned from some reference sequence(s).
Many thanks to Ben Kaehler for a proof of concept and mathematical assistance.
Disclaimer: This is unstable software at the moment. Use with caution.
pip install cython numpy nose pip install -U -e 'git+https://github.com/kdmurray91/mpg#egg=mpg'
The script mpg should have been installed. This learns the transition probablitlies from sequences (in Fasta format) and generates random sequences from this Markov chain.
To learn transition probablities for a 5th-order Markov chain from the Arabidopsis genome and store them in a file (ath.yml):
mpg -k 5 -d ath.yml TAIR_10.fasta
To use these transition probablitlies to generate a sequence of 1000bp:
mpg -l 1000 -r ath.yml
To do the above operations at once:
mpg -l 1000 -k 5 TAIR_10.fasta
TODO: Figure out how to actually get changelog content.
Changelog content for this version goes here.