Summary of Nebula: Neural Empirical Bayes Under Latent Representations For Efficient and Controllable Design Of Molecular Libraries, by Ewa M. Nowara et al.
NEBULA: Neural Empirical Bayes Under Latent Representations for Efficient and Controllable Design of Molecular Libraries
by Ewa M. Nowara, Pedro O. Pinheiro, Sai Pooja Mahajan, Omar Mahmood, Andrew Martin Watkins, Saeed Saremi, Michael Maser
First submitted to arxiv on: 3 Jul 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Biomolecules (q-bio.BM)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This research presents NEBULA, a latent 3D generative model for creating large molecular libraries around a seed compound. The goal is to efficiently generate high-quality samples while overcoming the limitations of existing methods like voxel-based approaches. NEBULA uses neural empirical Bayes sampling in a learned latent space from a vector-quantized variational autoencoder. This approach yields libraries nearly an order of magnitude faster than current methods without sacrificing quality, and it generalizes better to unseen drug-like molecules on two public datasets and multiple recently released drugs. The code is available online. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary NEBULA is a new way to make lots of chemical compounds quickly. It uses math to create these compounds from scratch, which helps scientists find new medicines. Right now, making lots of compounds takes too long, but NEBULA does it much faster without losing quality. This means scientists can test more ideas and maybe find better treatments. |
Keywords
» Artificial intelligence » Generative model » Latent space » Variational autoencoder
