Skip to content

Latest commit

 

History

History
112 lines (83 loc) · 3.26 KB

GettingStarted.md

File metadata and controls

112 lines (83 loc) · 3.26 KB
Raw

Getting Started

This guide walks you through using GRGL locally on a Linux or MacOS system. Requirements:

  • CMake v3.14 or newer
  • GCC or Clang that supports C++11
  • Python 3.7 or newer (and the development libraries -- e.g., python3-dev on Ubuntu)
  • zlib/libz development libraries

Build and Install

Create a Python virtual environment to work in.

python3 -m venv /path/to/anEnv

Replace /path/to/anEnv with any directory where you want to create the virtual environment.

Enable the Python virtual environment

source /path/to/anEnv/bin/activate

All Python packages you install will now be specific to this environment. You can deactivate the environment with the command deactivate. You can delete the environment directory when you are done with it.

Build and Install GRGL

git clone --recursive https://github.com/aprilweilab/grgl.git
cd grgl
pip install wheel
python setup.py bdist_wheel
pip install --force-reinstall dist/*.whl

Running an Example

Construct a GRG for one of the test inputs. From within the virtual environment, run:

grg construct test/endtoend/input/test-200-samples.vcf.gz

This will create the GRG test-200-samples.vcf.gz.final.grg in the directory you ran the command from. Using the Python API, you can load the GRG and calculate the allele counts using:

import pygrgl
from pygrgl import get_dfs_order, TraversalDirection

g = pygrgl.load_immutable_grg("test-200-samples.vcf.gz.final.grg")

node_sample_counts = {}
for node_id in get_dfs_order(g, TraversalDirection.DOWN, g.get_root_nodes()):
    count = 1 if g.is_sample(node_id) else 0
    for child_id in g.get_down_edges(node_id):
        count += node_sample_counts[child_id]
    node_sample_counts[node_id] = count

for node_id, mutation_id in g.get_node_mutation_pairs():
    mutation = g.get_mutation_by_id(mutation_id)
    if node_id != pygrgl.INVALID_NODE:
        count = node_sample_counts[node_id]
        print(f"Mutation({mutation_id}): ({mutation.position}, {mutation.allele}) has sample count {count}")

This code is also in the included Jupyter Notebook in grgl/jupyter/01.Intro.ipynb. To run that notebook, do the following in your grgl directory, in the virtual environment:

pip install jupyterlab # Install jupyter lab in your virtual environment if you haven't yet
jupyter lab .

In your browser, open the URL that JupyterLab prints to the screen. Navigate the tree to jupyter/01.Intro.ipynb and run the cells in the notebook one at a time.

Next steps

When working with anything but the smallest datasets, you'll want to convert your data to IGD before generating a GRG. This can be done with the grg convert command, which can convert any of .vcf, .vcf.gz, or .bgen to .igd.

Converting data to IGD

grg convert <.vcf/.vcf.gz/.bgen file> <.igd output filename>

Then GRG can be constructed from IGD.

grg construct <.igd filename>

Converting Tree-Sequence to GRG

grg convert <.trees input file> <.grg output filename>

Processing GRG file

  1. Allele frequency
grg process freq <.grg input filename>
  1. Graph statistics
grg process stats <.grg input filename>