splitp

Phylogenetic tools and methods involving splits and matrix rank


Keywords
python, parser, biology, linear-algebra, split, rank, phylogenetics, networks, trees, newick, newick-format, utas, splitp, flattenings, subflattenings, mathbio
License
MIT
Install
pip install splitp==0.1.7

Documentation

SplitP

Coverage Status

Python package which implements split- and rank-based tools for inferring phylogenies, such as flattenings and subflattenings.

Installation

The latest version of SplitP can be installed via the command pip install splitp

Examples

Import splitp and the associated helper functions

import splitp as sp
from splitp import tree_helper_functions as hf

Define trees and work with splits

splits = list(hf.all_splits(4))     # [01|23, 02|13, 03|12]
tree = sp.NXTree('((0,1),(2,3));')	
true_splits = tree.true_splits()    # 01|23

Let site patterns evolve under any submodel of the general markov model.

JC_subs_matrix = tree.build_JC_matrix(branch_length:=0.05)
tree.reassign_all_transition_matrices(JC_subs_matrix)
pattern_probs = tree.get_pattern_probabilities()
>             0         1
      0    AAAA  0.185844
      1    AAAC  0.003262
      ..    ...       ...
      254  TTTG  0.003262
      255  TTTT  0.185844

Simulate sequence alignments from pattern distributions

pattern_frequencies = tree.draw_from_multinomial(pattern_probs, 100)
>         0    1
    0  AAAA  0.22
    1  AAAC  0.01
    ..  ...   ...
    2  CCGC  0.03
    3  TTTT  0.14

Reconstruct trees using split based methods including flattenings:

F1 = tree.flattening('01|23', pattern_frequencies)
F2 = tree.flattening('02|13', pattern_frequencies)
print(tree.split_score(F1) < tree.split_score(F2))    # True

Or subflattenings:

SF = tree.signed_sum_subflattening('01|23', pattern_probs)
print(tree.split_score(SF))   # 0.0

For more functionality please see the documentation at splitp.joshuastevenson.me.

Please see CONTRIBUTING.md for information on contributing to this project.