This repository contains software, data, and results for the BUSTED[S]-MH manuscript.

About

A link for our preprint version will be provided here.

Conventional models may fail to accurately depict the evolutionary pressures and constraints at individual sites and along branches for a protein-coding sequence. We have previously demonstrated widespread empirical support for models which account for instantaneous multiple nucleotide changes within a codon (MH). Here, we present an extension of the commonly used branch-site framework of BUSTED[S], a method for the inference of gene-wide episodic diversifying selection, developed in order to account for the impact of MH events (BUSTED[S]-MH). Our new model contributes towards an increased interest in the advances our understanding of biological realism within codon models and offers an attractive alternative to traditional models used for genome-wide scans of adaptive evolution. Comprehensive analysis of the effects of MH mutations which may act on a subset of branches in a phylogeny and at a subset of sites within the gene may shed light on the additional pathways available for genes to embark upon during the evolutionary process.

Requirements

1. HyPhy
2. Snakemake

Installation (Conda-based)

1. git clone https://github.com/aglucaci/BUSTEDS-MH.git
2. cd BUSTEDS-MH
3. conda env create -f environment.yml.
   This will create a conda environment called (BUSTEDS-MH) with the necessary dependencies.
4. At this point, run conda activate BUSTEDS-MH and your environment will be ready to go.
5. We also require the HyPhy-analyses repository for the BUSTED[S]-MH model batch file, you can clone this repository within the working direcory with `git clone https://github.com/veg/hyphy-analyses

Datasets

We use a number of small empirical datasets, those available are available here, Empirical 14 datasets, and described in detail in the manuscript.

We use a number of large empirical datasets, those available are below, and described in detail in the manuscript.

1. [Selectome, Euteleostomi, Version 6](https://github.com/aglucaci/BUSTEDS-MH/tree/develop/data/Empirical_Selectome)

We use a number of simulated datasets, those available are below, and described in detail in the manuscript.

1. [Simulated 16/31 sequences, Power simulations](https://github.com/aglucaci/BUSTEDS-MH/tree/develop/data/SIMULATED_16_31)
2. [Null Simulations](https://github.com/aglucaci/BUSTEDS-MH/tree/develop/data/SIMULATED_null)

Minimum Configuration

Everything is controlled through the cluster.json and config.yml files.

The cluster JSON file contains information about where to submit your jobs on a HPC system.

The config.yml specifies the following variables: DataDirectory, the full path to your data directory SubDirectory, the subdirectory of that folder to analyze FileEnding, the file ending to search for, for example .fasta or .nex OutputDirectory, the ouput directory for your HyPhy JSON results BUSTEDSMHBF, the full path to your BUSTED[S]-MH HyPhy Batch file, typically this is located within ./hyphy-analyses/BUSTED-MH/BUSTED-MH.bf