_images/logo_condensed.png

OrthoHMM

OrthoHMM infers orthogroups using a built-in profile-HMM search engine and Leiden clustering. The standard pipeline does not require external bioinformatics executables. Optional modes support high-sensitivity profile expansion, phylogenetic reconciliation, and legacy HMMER/MCL inference.

Installation

Use Python 3.10 or newer and an isolated environment. These commands install the development application from source, not a potentially older PyPI release.

git clone --depth 1 https://github.com/JLSteenwyk/orthohmm.git
cd orthohmm
python3 -m venv .venv
source .venv/bin/activate
python -m pip install .
orthohmm --help

A C compiler with OpenMP support builds the native CPU kernels during installation. On Linux, install GCC and its development tools. On macOS, install Homebrew GCC and ensure gcc resolves to that compiler when installing. Standard search can fall back to Numba when compilation is unavailable; high-sensitivity MSA profile expansion requires native kernels. Set ORTHOHMM_CPU_TARGET=baseline during installation when building for compatible machines other than the build host.

Usage

Place one protein FASTA file per species in an input directory and create the output directories before running.

mkdir results sensitive_results
orthohmm proteomes/ -o results/ -c 8
orthohmm proteomes/ -o sensitive_results/ -c 8 --accuracy_profile high_sensitivity

See Advanced Usage for options and output files.

Optional dependencies

  • Phylogenetic reconciliation: install '.[phylogeny]' and provide MAFFT and FastTree executables. A supplied or inferred species tree can be used.

  • Legacy sequence search: install HMMER and select --search_mode phmmer.

  • Legacy clustering: install MCL and select --clustering mcl.

  • Python HMMER control helpers: install '.[hmmer]' for their Biopython dependency. This extra is not needed by the built-in search engine.

Research and citation

The application repository keeps source, tests, and user documentation. Manuscripts, benchmark datasets, campaign scripts, and historical results are preserved on the research-archive-20261029 branch. See Research and Evaluation for the research snapshot and evaluation limitations.

Please cite OrthoHMM: Improved Inference of Ortholog Groups using Hidden Markov Models, Steenwyk et al. (2024), bioRxiv, doi:10.1101/2024.12.07.627370. Citation metadata are also available in CITATION.cff.