BERTology Meets Biology: Interpreting Attention in Protein Language Models

This repository is the official implementation of BERTology Meets Biology: Interpreting Attention in Protein Language Models.

ProVis Attention Visualizer

Installation

General requirements: * Python >= 3.6 * PyTorch (See installation instructions here.)

Specific libraries:

pip install biopython==1.77
pip install tape-proteins==0.4

NGLViewer (based on instructions found here):

• Available on

  conda-forge

  channel

  conda install nglview -c conda-forge
  jupyter-nbextension enable nglview --py --sys-prefix
  

# if you already installed nglview, you can

upgrade

• Available on PyPI

   pip install nglview
   jupyter-nbextension enable nglview --py --sys-prefix

To use in Jupyter Lab you need to install appropriate extension:

jupyter labextension install  nglview-js-widgets

Execution

cd /notebooks
jupyter notebook provis.ipynb

You may edit the notebook to choose other proteins, attention heads, etc. The visualization tool is based on the excellent nglview library.

Experiments

Installation

cd 
python setup.py develop

Datasets

To download additional required datasets from TAPE:

cd /data
wget http://s3.amazonaws.com/proteindata/data_pytorch/secondary_structure.tar.gz
tar -xvf secondary_structure.tar.gz && rm secondary_structure.tar.gz
wget http://s3.amazonaws.com/proteindata/data_pytorch/proteinnet.tar.gz
tar -xvf proteinnet.tar.gz && rm proteinnet.tar.gz

Attention Analysis

The following steps will recreate the reports currently found in /reports/attention_analysis

Before performing steps, navigate to appropriate directory:

cd /protein_attention/attention_analysis

Amino Acids

Run analysis (may wish to run in background):

sh scripts/compute_aa_features.sh

Run report from extract file:

python report_edge_features.py edge_features_aa
python report_aa_correlations.py edge_features_aa

Secondary Structure

Run analysis:

sh scripts/compute_sec_features.sh

Run reports:

python report_edge_features.py edge_features_sec

Contact Maps

Run analysis:

sh scripts/compute_contact_features.sh

Run report:

python report_edge_features.py edge_features_contact

Binding Sites

Run analysis:

sh scripts/compute_site_features.sh

Run report:

python report_edge_features.py edge_features_sites

Combined features

Create report of all features combined

python report_edge_features_combined.py edge_features_aa edge_features_sec edge_features_contact edge_features_sites

Probing Analysis

The following steps will recreate the reports currently found in /reports/probing

Navigate to directory:

cd /protein_attention/probing

Training

Train diagnostic classifiers. Each script will write out an extract file with evaluation results. Note: each of these scripts may run for several hours.

sh scripts/probe_ss4_0_all
sh scripts/probe_ss4_1_all
sh scripts/probe_ss4_2_all
sh scripts/probe_sites.sh
sh scripts/probe_contacts.sh

Reports

python report.py

License

This project is licensed under BSD3 License - see the LICENSE file for details

Acknowledgments

This project incorporates code from the following repo: * https://github.com/songlab-cal/tape

Citation

When referencing this repository, please cite this paper.

@misc{vig2020bertology,
    title={BERTology Meets Biology: Interpreting Attention in Protein Language Models},
    author={Jesse Vig and Ali Madani and Lav R. Varshney and Caiming Xiong and Richard Socher and Nazneen Fatema Rajani},
    year={2020},
    eprint={2006.15222},
    archivePrefix={arXiv},
    primaryClass={cs.CL},
    url={https://arxiv.org/abs/2006.15222}
}