# Are VAEs Bad at Reconstructing Molecular Graphs?

arXiv: https://arxiv.org/abs/2305.03041 · HTML: https://mlsanity.com/paper/2305.03041/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 34 references that are in the map.
- First posted on arXiv 4 May 2023.

## Builds on (most cited first, 17 of 17)

- [Automatic chemical design using a data-driven continuous representation of molecules](https://mlsanity.com/paper/1610.02415/) (2016)
- [Recipe for a General, Powerful, Scalable Graph Transformer](https://mlsanity.com/paper/2205.12454/) (2022)
- [Junction Tree Variational Autoencoder for Molecular Graph Generation](https://mlsanity.com/paper/1802.04364/) (2018)
- [GuacaMol: Benchmarking Models for De Novo Molecular Design](https://mlsanity.com/paper/1811.09621/) (2018)
- [Hierarchical Generation of Molecular Graphs using Structural Motifs](https://mlsanity.com/paper/2002.03230/) (2020)
- [Fr'echet ChemNet Distance: A metric for generative models for molecules in drug discovery](https://mlsanity.com/paper/1803.09518/) (2018)
- [Molecular representation learning with language models and domain-relevant auxiliary tasks](https://mlsanity.com/paper/2011.13230/) (2020)
- [Molecule Attention Transformer](https://mlsanity.com/paper/2002.08264/) (2020)
- [Learning to Extend Molecular Scaffolds with Structural Motifs](https://mlsanity.com/paper/2103.03864/) (2021)
- [Hit and Lead Discovery with Explorative RL and Fragment-based Molecule Generation](https://mlsanity.com/paper/2110.01219/) (2021)
- [Message-passing neural networks for high-throughput polymer screening](https://mlsanity.com/paper/1807.10363/) (2018)
- [Dual-view Molecule Pre-training](https://mlsanity.com/paper/2106.10234/) (2021)
- [Relative Molecule Self-Attention Transformer](https://mlsanity.com/paper/2110.05841/) (2021)
- [C5T5: Controllable Generation of Organic Molecules with Transformers](https://mlsanity.com/paper/2108.10307/) (2021)
- [DeepScaffold: a comprehensive tool for scaffold-based de novo drug discovery using deep learning](https://mlsanity.com/paper/1908.07209/) (2019)
- [Generating Focussed Molecule Libraries for Drug Discovery with Recurrent Neural Networks](https://mlsanity.com/paper/1701.01329/) (2017)
- [Scaffold-based molecular design using graph generative model](https://mlsanity.com/paper/1905.13639/) (2019)

## Built on by (most cited first, 1 of 1)

- [Auto-encoding Molecules: Graph-Matching Capabilities Matter](https://mlsanity.com/paper/2503.00426/) (2025)

## Abstract (arXiv)

Many contemporary generative models of molecules are variational auto-encoders of molecular graphs. One term in their training loss pertains to reconstructing the input, yet reconstruction capabilities of state-of-the-art models have not yet been thoroughly compared on a large and chemically diverse dataset. In this work, we show that when several state-of-the-art generative models are evaluated under the same conditions, their reconstruction accuracy is surprisingly low, worse than what was previously reported on seemingly harder datasets. However, we show that improving reconstruction does not directly lead to better sampling or optimization performance. Failed reconstructions from the MoLeR model are usually similar to the inputs, assembling the same motifs in a different way, and possess similar chemical properties such as solubility. Finally, we show that the input molecule and its failed reconstruction are usually mapped by the different encoders to statistically distinguishable posterior distributions, hinting that posterior collapse may not fully explain why VAEs are bad at reconstructing molecular graphs.