Skip to content
Go to file

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time


A massively parallel tool for model selection and tree inference on thousands of genes


ParGenes is for you if:

  • you have several MSAs (typically gene alignments).
  • you want to run maximum likelihood tree inference (RAxML) independently on each of them. For instance, to get one gene tree per gene alignment.
  • you want to run these jobs in parallel, (single or multiple nodes).

In addition, ParGenes:

  • can find the best-fit model with ModelTest and use this model in the RAxML calls.
  • has a checkpoint mechanism.
  • filters out and reports a list of the MSAs that RAxML can not process.
  • handles multiple starting trees, bootstrap replicates, support value. ParGenes can run these searches simultaneously, and thus improves RAxML parallelization scheme.
  • provides a (global or per-MSA) way to customize the modeltest and RAxML calls.
  • can infer the optimal number of cores to assign to a given ParGenes call.


  • A linux or MacOS platform
  • gcc 5.0 or > (we did not try with clang yet)
  • CMake 3.6 or >
  • Either MPI or OpenMP. MPI for multiple nodes parallelization.


Please use git, and clone with --recursive!!!

git clone --recursive

To build the sources:


To parallelize the compilation with 10 cores:

./ 10

Updating the repository

Instead of using: git pull please use: ./

Rational: we use git submodule, and git pull might not be enough to update all the changes. The will update all the changes properly.


See the wiki (

Documentation and Support

Documentation: in the github wiki.

Also please check the online help with python3 pargenes/ --help

A suggestion, a bug to report, a question? Please use the RAxML google group.


Before citing ParGenes, please make sure you read


A massively parallel tool for model selection and tree inference on thousands of genes




No packages published
You can’t perform that action at this time.