Skip to content

matsengrp/historydag

Repository files navigation

HistoryDAG

This package provides an implementation for a history DAG object, which compactly expresses a collection of internally labeled trees which share a common set of leaf labels.

Getting Started

HistoryDAG is on PyPI! Install with pip install historydag.

Alternatively, clone the repo and perform pip install -e historydag.

The most common input for DAG construction is collections of ete3 phylogenetic trees with full sequences at internal nodes stored in the sequence attribute. There is sample data like this in sample_data/. For example:

import historydag as hdag
import pickle

with open('sample_data/toy_trees.p', 'rb') as fh:
    ete_trees = pickle.load(fh)

# Build the DAG, specifying to only use the `sequence` attribute for node
# labels (in general, one could use other attributes as well).
dag = hdag.history_dag_from_etes(ete_trees, ['sequence'])
dag.count_histories()  # 1041

# "Complete" the DAG, adding all allowable edges.
dag.make_complete()
dag.count_histories()  # 3431531

# Show counts of trees with various parsimony scores.
dag.hamming_parsimony_count()

# "Trim" the DAG to make it only display minimum-weight trees.
dag.trim_optimal_weight()
# With default args, same as hamming_parsimony_count
dag.weight_count()  # Counter({75: 45983})

# "Collapse" the DAG, contracting zero-weight edges.
dag.convert_to_collapsed()

dag.weight_count()  # Counter({75: 1208})
dag.count_topologies()  # 1054 unique topologies, ignoring internal labels

# To count parsimony score and the number of unique nodes in each tree jointly:
node_count_funcs = hdag.utils.AddFuncDict(
    {
        "start_func": lambda n: 0,
        "edge_weight_func": lambda n1, n2: n1.label != n2.label,
        "accum_func": sum,
    },
    name="NodeCount",
)
dag.weight_count(**(node_count_funcs + hdag.utils.hamming_distance_countfuncs))
# Counter({(50, 75): 444, (51, 75): 328, (49, 75): 270, (52, 75): 94, (48, 75): 68, (53, 75): 4})

# To trim to only the trees with 48 unique node labels:
dag.trim_optimal_weight(**node_count_funcs, optimal_func=min)

# Sample a tree from the dag and make it an ete tree.
t = dag.sample().to_ete()

# the history DAG also supports indexing and iterating:
t = dag[0].to_ete()
trees = [tree for tree in dag]

# Another method for fetching all trees in the dag is provided, but the order
# will not match index order:
scrambled_trees = list(dag.get_histories())

# Union is implemented as dag merging, including with sequences of dags.
newdag = dag[0] | dag[1]
newdag = dag[0] | (dag[i] for i in range(3,5))

Highlights

  • History DAGs can be created with top-level functions like
    • from_newick
    • from_ete
    • history_dag_from_newicks
    • history_dag_from_etes
  • Trees can be extracted from the history DAG with methods like
    • HistoryDag.get_histories
    • HistoryDag.sample
    • HistoryDag.to_ete
    • HistoryDag.to_newick and HistoryDag.to_newicks
  • Simple history DAGs can be inspected with HistoryDag.to_graphviz
  • The DAG can be trimmed according to arbitrary tree weight functions. Use HistoryDag.trim_optimal_weight.
  • Disambiguation of sparse ambiguous labels can be done efficiently, but doesn't scale well. Use HistoryDag.explode_nodes followed by HistoryDag.trim_optimal_weight.
  • Weights of trees in the DAG can be counted, according to arbitrary weight functions using HistoryDag.weight_count. The class utils.AddFuncDict is provided to manage these function arguments, and implements addition so that different weights can be counted jointly. These same functions can be used in trimming.

Important Details

In order to create a history DAG from a collection of trees, each tree should meet the following criteria:

  • No unifurcations, including at the root node. Each node must have at least two children, unless it's a leaf node.
  • The label attributes used to construct history DAG labels must be unique, because history DAG nodes which represent leaves must be labeled uniquely.

Documentation

Docs are available at https://matsengrp.github.io/historydag.

To build docs, after installing requirements from requirements.txt, do make docs to build sphinx documentation locally. You'll find it at docs/_build/html/index.html.