RekerLab/mgktools

Package for marginalized graph kernel.

mgktools

Python Package using marginalized graph kernel (MGK) to predict molecular properties.

Installation

Suggested Package Versions: Python==3.10, GCC==11.2, CUDA==11.7.

pip install numpy==1.22.3 git+https://gitlab.com/Xiangyan93/graphdot.git@feature/xy git+https://github.com/bp-kelley/descriptastorus
pip install mgktools

Usage

See notebooks

Hyperparameters

hyperparameters contains the JSON files that define the hyperparameters for MGK.

Related work

• Predicting Single-Substance Phase Diagrams: A Kernel Approach on Graph Representations of Molecules
• A Comparative Study of Marginalized Graph Kernel and Message-Passing Neural Network
• Interpretable Molecular Property Predictions Using Marginalized Graph Kernels