Python libraries and utilities for computational biology.

This package contains algorithms related to several areas of genomics, phylogenetics, and population genetics. Some of the highlights include:

• reading, writing, and manipulating phylogenetic trees
• reconciling gene-trees with species-trees
• inferring gene duplications, losses, and horizontal transfers
• methods for coalescent processes, incomplete lineage sorting
• methods for ancestral recombination graphs (ARGs)
• finding syntenic regions (i.e. co-linear orthology) between genomes
• processing common file formats: FASTA, PHYLIP, newick, nexus, etc.

In addition to computational biology-specific methods, this package also contains general utilities for working with scientific data:

• sparse matrix file formats
• reading, writing, and manipulating tables of data
• working with intervals (e.g. intersection, union, etc)
• plotting (Gnuplot, Rpy)
• statistics
• general data-structures and algorithms: quad trees, Union-Find, HHMs, clustering

The compbio package is available for download from several sources:

• https://github.com/mdrasmus/compbio
• https://pypi.python.org/pypi/compbio

Most modules in this package can be used without any additional dependencies.

For plotting modules, the dependencies include:

• gnuplot
• R

For some scientific methods, the dependencies include:

• scipy

For development of the compbio package itself, dependencies can be installed with pip:

pip install -r requirements-dev.txt

The compbio package is available on pypi, and can be installed using pip:

pip install compbio

These packages can be installed from the source directory using:

python setup.py install

Optionally, the libraries can be used directly from the source directory by configuring one's environment variables as follows (assuming bash shell):

export PATH=$PATH:path/to/compbio/bin
export PYTHONPATH=$PYTHONPATH:path/to/compbio

These libraries were built up over the course of the Ph.D. of the author, Matthew D. Rasmussen (http://mattrasmus.com, [email protected]). Many of the methods here were utilized in several published software projects including:

• ARGweaver: Rasmussen, Siepel. Genome-wide inference of ancestral recombination graphs. ArXiv. 2013.
• DLCoal: Rasmussen, Kellis. Unified modeling of gene duplication, loss, and coalescence using a locus tree. Genome Research. 2012.
• SPIMAP: Rasmussen, Kellis. A Bayesian approach for fast and accurate gene tree reconstruction. Molecular Biology and Evolution. 2010.
• SPIDIR: Rasmussen, Kellis. Accurate gene-tree reconstruction by learning gene- and species-specific substitution rates across multiple complete genomes. Genome Research. 2007.

Minor note: Although the libraries of this package supports each of these software packages, this package is not a required dependency. Instead each software package contains its own private copy of modules taken from this package.