Replication of the Principal Odor Map paper by Brian K. Lee et al. (2023) [1]. The model is implemented such that it integrates with DeepChem.
| Model | Data | Type | ROC-AUC Score |
|---|---|---|---|
| MPNNPOMModel | curated_GS_LF_merged_4983.csv | 5-Fold CV with ensemble of 10 models per fold | 0.8872 |
-
pip install openpom
openpom requires cuda verion of dgl libraries
For cuda 11.7, steps below can be followed:
-
pip install dgl==1.1.1 -f https://data.dgl.ai/wheels/cu117/repo.html
Note: If you are using Amazon Linux 2 based OS on EC2 instance, use:
pip install dgl==1.1.2 -f https://data.dgl.ai/wheels/cu117/repo.html- Fork the OpenPOM repository and clone the forked repository
git clone https://github.com/YOUR-USERNAME/openpom.git
cd openpom- Setup conda environment
conda create -n open_pom python=3.9
conda activate open_pom- Install openpom
pip install .or (for developing)
python setup.py develop- Install DGL cuda libs
pip install dgl==1.1.1 -f https://data.dgl.ai/wheels/cu117/repo.htmlExample notebooks for model training and finetuning are available here.
Aryan Amit Barsainyan, National Institute of Technology Karnataka, India: code, data cleaning, model development
Ritesh Kumar, CSIR-CSIO, Chandigarh, India: data cleaning, hyperparameter optimisation
Pinaki Saha, University of Hertfordshire, UK: discussions and feedback
Michael Schmuker, University of Hertfordshire, UK: conceptualisation, project lead
[1] A Principal Odor Map Unifies Diverse Tasks in Human Olfactory Perception.
Brian K. Lee, Emily J. Mayhew, Benjamin Sanchez-Lengeling, Jennifer N. Wei, Wesley W. Qian, Kelsie A. Little, Matthew Andres, Britney B. Nguyen, Theresa Moloy, Jacob Yasonik, Jane K. Parker, Richard C. Gerkin, Joel D. Mainland, Alexander B. Wiltschko
Science381,999-1006(2023).DOI: 10.1126/science.ade4401
bioRxiv 2022.09.01.504602; doi: https://doi.org/10.1101/2022.09.01.504602