GraphNVP: An Invertible Flow Model for Generating Molecular Graphs
This work addresses molecular graph generation for drug discovery and materials science, representing an incremental advancement by applying invertible flows to this domain.
The authors tackled the problem of generating molecular graphs by introducing GraphNVP, an invertible flow model that decomposes graph generation into adjacency tensor and node attribute steps, resulting in efficient generation of valid molecules with minimal duplication and the ability to produce molecules with desired chemical properties.
We propose GraphNVP, the first invertible, normalizing flow-based molecular graph generation model. We decompose the generation of a graph into two steps: generation of (i) an adjacency tensor and (ii) node attributes. This decomposition yields the exact likelihood maximization on graph-structured data, combined with two novel reversible flows. We empirically demonstrate that our model efficiently generates valid molecular graphs with almost no duplicated molecules. In addition, we observe that the learned latent space can be used to generate molecules with desired chemical properties.