Copyright (C) 2020 Arthur Zwaenepoel

This repository is not actively maintained, use at own risk

py-adhore

Utilities for doing and visualizing synteny/co-linearity analyses using I-ADHoRe. This should eventually replace the tools provided by wgd syn in the wgd package. It is mainly a wrapper to enable easier usage of the very fast and sensitive I-ADHoRe program for synteny and co- linearity inference in large comparative genomic data sets; and provides the tools for associated visualizations one usually desires.

To install, open a terminal session, clone the repository, cd into it and type

$ pip install .

Works with python3 (you may need to use pip3 instead of pip in the above command; consider using a virtualenv).

I-ADHoRe can be downloaded here (If the page prompts you to go to the updated website, decline and go to the bottom of the page to download the package (v3.0)).

Usage

Running I-ADHoRe on OrthoFinder gene family data

The example directory of this repository contains an example data set, consisting of gene families (og.csv) inferred with OrthoFinder for Arabidopsis thaliana and Vitis vinifera and two .gff files for the associated genes of these species (downloaded, along with the proteome fasta files used in OrthoFinder, from PLAZA 4.0).

To run I-ADHoRe for this data set you can use the following command

$ py-adhore of ./Orthogroups.tsv ath,vvi ./ath.gff ./vvi.gff -f gene -a Name -n 4 --run

This will use four threads (-n 4) for the I-ADHoRe algorithm. Say you only wanted to do a within-genome comparison for Vitis, you could simply do

$ py-adhore of ./Orthogroups.tsv vvi ./vvi.gff -f gene -a Name -o vvi_out -n 4 --run

You can find more options and instructions using py-adhore -of --help.

To get the associated synteny clusters, use:

$ py-adhore cl py-adhore.out/genes_data.csv py-adhore.out/i-adhore-out/anchorpoints.txt

We make the following notes:

1. It is very important that the gene IDs in the OrthoFinder gene families correspond to the gene IDs found in the features of the gff file that correspond with the -f setting (here gene features) at the attribute specified with the -a option (here Name). If this is not clear have a look at the example data.
2. One can use any number and any subset of species from the gene families file.
3. One can use a plain MCL gene families file (where each line is whitespace separated list of genes) by using the --mcl flag. In this case, gene IDs must have species IDs as prefixes
4. One can choose to just generate the configuration files for I-ADHoRe without running it, simply omit the --run flag.
5. Note that if you're using OrthoFinder, you may want to add the singleton genes in Orthogroups.UnassignedGenes.tsv to the Orthogroups.tsv file before using with py-adhore, otherwise you will ignore all genes that are in singleton gene families.

Visualization

One can visualize the results in a circos diagram using Circos or Circos.js as follows (example for circos.js):

$ py-adhore cc --js py-adhore.out/i-adhore-out/segments.txt ./py-adhore.out/genes_data.csv

This will create a directory py-adhore.out/circos with a HTMl file in the case of Circos.js. If the --js flag is omitted configuration files for Circos are generated.

Citation

If you use this code, please do not forget to cite

Proost, Sebastian, et al. "i-ADHoRe 3.0—fast and sensitive detection of genomic homology in extremely large data sets."
Nucleic acids research 40.2 (2011): e11-e11.