https://github.com/4ment/dodonaphy-experiments

https://github.com/4ment/dodonaphy-experiments

Last synced: 6 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/4ment/dodonaphy-experiments
• Owner: 4ment
• Created: 2024-02-24T21:42:38.000Z (over 2 years ago)
• Default Branch: main
• Last Pushed: 2024-03-30T23:53:29.000Z (about 2 years ago)
• Last Synced: 2025-05-17T05:09:13.500Z (about 1 year ago)
• Language: Python
• Size: 79.1 KB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# dodonaphy-experiments

[![Docker Image CI](https://github.com/4ment/dodonaphy-experiments/actions/workflows/docker-image.yml/badge.svg)](https://github.com/4ment/dodonaphy-experiments/actions/workflows/docker-image.yml)

This repository contains the pipeline and data sets supporting the results of the following article:

Macaulay M and Fourment M. Differentiable Phylogenetics via Hyperbolic Embeddings with Dodonaphy. [arXiv:2309.11732](https://arxiv.org/abs/2309.11732)

## Dependencies

You will need to install [nextflow](https://www.nextflow.io) to run the pipeline. For ease of use the pipeline can be set up with either the package manager [conda](https://conda.io) or [docker](https://www.docker.com). Singularity is another option for running the pipeline on an HPC.

## Download pipeline

    git clone http://github.com/4ment/dodonaphy-experiments.git

    cd dodonaphy-experiments/

    chmod +x bin/*.py

## Pipeline without docker/singularity

### Installing dependencies with conda

    conda env create -f environment.yml

    conda activate dodonaphy

    git clone http://github.com/mattapow/hydraPlus.git

    pip install hydraPlus/

    git clone http://github.com/mattapow/dodonaphy.git

    pip install dodonaphy/

### Running the pipeline

    nextflow run main.nf

## Pipeline with docker or singularity

There is no need to install dependencies with docker or singularity.

### Running the pipeline with docker

    nextflow run main.nf -profile docker

### Running the pipeline with singularity and PBS

    nextflow -C configs/uts.config run main.nf -profile singularity

Since the pipeline will take weeks to run to completion one should use a high performance computer. An example of a configuration file for pbspro can be found in the [configs](configs/) folder.

## Generating figures

Table 1: Comparison of marginal log-likelihood estimates.

    find results/vi -name samples.t|grep b1|xargs python bin/marginal.py

Figure 2: Difference in maximum log likelihood estimates compared to RAxML across all datasets DS1-8.

    find results/hmap -name "*.csv"| xargs python bin/ml_performance.py

Figure 3: Variational approximation in $H^3$ compared to MCMC.

    python bin/split_lengths_prepare.py --output branch_lengths_by_treelist_sorted.csv results/vi/DS1/vi/up_nj/d3_lr2_i3_b3/samples.t results/mrbayes/DS1/DS1.nex.run*.t

    python bin/split_lengths_plot.py branch_lengths_by_treelist_sorted.csv

Figure 4: Effect of the number of boosts on the final SIWAE estimate for DS1.

    find results/vi/DS1 -name vi.log| xargs python bin/SIWAE_boosts.py