PyASR

Ancestral Sequence Reconstruction in Python

PyASR offers a modern Python interface to PAML (“Phylogenetic Analysis by Maximum Likelihood”)—tuned specifically for reconstruction
ancestral protein/DNA sequences.

NOTE: PyASR currently only supports protein reconstructions. This is a work in progress.

Basic Usage

import phylopandas as pd
import dendropy as d
import pyasr
# Use phylopandas to read a set of ancestor.s
df_seqs = pd.read_fasta('test.fasta')
# Use dendropy to read in tree.
tree = d.Tree.get(path='tree.newick', schema='newick')
# Reconstruct nodes in tree.
tree, df_seqs, df_anc = pyasr.reconstruct(df_seqs, tree, working_dir='test', alpha=1.235)
# Write out ancestor dataframe to a CSV file.
df_anc.to_csv('ancestors.csv')

We can visualize the ancestors side-by-side with the tree using inside of JupyterLab
thanks to the ToyTree library.

Install

This package is released on PyPi. You can install using pip:

pip install pyasr

To get the development version:

git clone
cd
pip install -e .

Dependencies

The actual reconstruction calculation are done using PAML. This requires PAML to be
installed and the codeml/baseml executables exported to your $PATH environment variable. Directions for installing PAML can be found on the PAML website.

The following Python dependencies are required for PyASR to work.

• Pandas
• Biopython
• PhyloPandas
• DendroPy