damiano-sg/phylogeny-msa-transformer

Supporting repository for "Protein language models trained on multiple sequence alignments learn phylogenetic relationships" (preprint: https://doi.org/10.1101/2022.03.29.486219)

Phylogeny-MSA-Transformer

Supporting repository for "Protein language models trained on multiple sequence alignments learn phylogenetic relationships" (Lupo, Sgarbossa, and Bitbol, 2022). The MSA Transformer model used here was introduced in (Rao el al, 2021).

Getting started

Clone this repository on your local machine by running

git clone [email protected]:Bitbol-Lab/Phylogeny-MSA-Transformer.git

and move inside the root folder. We recommend creating and activating a dedicated conda or virtualenv Python virtual environment. Then, install the required libraries:

python -m pip install -U -r requirements.txt

Requirements

In order to run the notebooks, the following python packages are required:

• tqdm
• jupyter
• matplotlib
• statsmodels
• biopython
• swalign
• esm==0.4.0

prody and HMMER are required to run the Python script data/Pfam_Seed/fetch_seed_MSAs.py, if you wish to create new Pfam full MSAs instead of using the ones provided.

Citation

Our work can be cited using the following BibTeX entry:

@misc{lupo2022protein,
      title={Protein language models trained on multiple sequence alignments learn phylogenetic relationships},
      author={Lupo, Umberto and Sgarbossa, Damiano and Bitbol, Anne-Florence},
      year={2022},
      doi={10.1101/2022.03.29.486219},
      url={https://www.biorxiv.org/content/early/2022/03/30/2022.03.29.486219},
      eprint={https://www.biorxiv.org/content/early/2022/03/30/2022.03.29.486219.full.pdf},
      journal={bioRxiv}
}