Accelerating Antimicrobial Discovery with Controllable Deep Generative Models and Molecular Dynamics

This work will be published in Nature Biomedical Engineering on March 11, 2021

URL : https://www.nature.com/articles/s41551-021-00689-x

De novo therapeutic design is challenged by a vast chemical repertoire and multiple constraints, e.g., high broad-spectrum potency and low toxicity. This project proposes CLaSS (Controlled Latent attribute Space Sampling) - an efficient computational method for attribute-controlled generation of molecules, which leverages guidance from classifiers trained on an informative latent space of molecules modeled using a deep generative autoencoder. We screen the generated molecules for additional key attributes by using deep learning classifiers in conjunction with novel features derived from atomistic simulations.

Setup

• The amp_gen.yml lists are the required dependencies for the project.
• Use amp_gen.yml to create your own conda environment to run this project. Command: conda-env create -f amp_gen.yml

Usage

Phase 1: Autoencoder (VAE/WAE) Training

• ./run.sh. This will run with default config from cfg.py. Since cfg.runname=default the output goes to output/default and tb/default.
• python main.py --tiny 1 for fast testing with default config file.
• Additionally, one could explicitly run the individual scripts as follows:

  • python main.py --phase 1

  • python static_eval.py --config_json output/dir/config_overrides.json

Phase 2: CLaSS (Controlled Latent attribute Space Sampling)

• python sample_pipeline.py --config_json output/default/config_overrides.json --samples_outfn_prefix samples --Q_select_amppos 0

Data:

• data_processing/data dir has the short versions of data files required by our data curation code data_processing/create_datasets.py
• For the full version of dataset use following links to download full version of data files that are publicly available.
• UNIPROT: [https://www.uniprot.org/uniprot/?query=reviewed:yes] and [https://www.uniprot.org/uniprot/?query=reviewed:no]
• SATPDB: [http://crdd.osdd.net/raghava/satpdb/]
• DBAASP: [https://dbaasp.org]
• AMPEP: [https://cbbio.cis.um.edu.mo/software/AmPEP/]
• ToxinPred: [https://webs.iiitd.edu.in/raghava/toxinpred/dataset.php]

Related Visualization Tools

• Peptide Walker : https://peptide-walk.mybluemix.net
• Cogmol Drug Exploration: https://covid19-mol.mybluemix.net

Citations

Please cite the following articles:

@article{das2020accelerating,
  title={Accelerating Antimicrobial Discovery with Controllable Deep Generative Models and Molecular Dynamics},
  author={Das, Payel and Sercu, Tom and Wadhawan, Kahini and Padhi, Inkit and Gehrmann, Sebastian and Cipcigan, Flaviu and Chenthamarakshan, Vijil and Strobelt, Hendrik and Santos, Cicero dos and Chen, Pin-Yu and others},
  journal={arXiv preprint arXiv:2005.11248},
  year={2020}
}

@article{chenthamarakshan2020cogmol,
  title={CogMol: Target-specific and selective drug design for COVID-19 using deep generative models},
  author={Chenthamarakshan, Vijil and Das, Payel and Hoffman, Samuel C and Strobelt, Hendrik and Padhi, Inkit and Lim, KW and others},
  journal={arXiv: 2004.01215},
  year={2020}
  }