Marta Sales-Pardo & Roger Guimerà . Complex systems & networks; Statiscal learning; Comput. social science; Systems biology at @universitatURV @icreacommunity
SEES Lab
Loading...
Because chemical validity is satisfied by construction, #CoCoGraph needs much fewer parameters than the state of the art and it can focus on learning subtle chemical patterns, leading to generation of molecules that are more realistic than existing models
102 organic chemists shown real vs AI-generated molecules. Couldn't tell them apart — 62% accuracy ≈ random. That's CoCoGraph: graph diffusion with 100% chemical validity, 534K params vs 4.6M for comparable models. Constraints in the math, not the model.
www.nature.com/articles/s42...
A new AI tool can generate millions of chemically valid molecules, offering a faster and more efficient approach to exploring potential compounds for drug development and materials science. doi.org/hb2mwd
A collaborative constrained graph diffusion model for the generation of realistic synthetic molecules [new]
...that ensures generated novel molecules are valid and highly realistic.
To verify this, we implemented and analyzed the results of a Turing-like test, where experts were asked to distinguish between real and generated molecules
#CoCoGraph is a collaborative constrained graph diffusion model for the generation of realistic synthetic molecules. Unlike previous generative models for molecules, we use ideas from graph theory to guarantee that all generated molecules are chemically valid
Can we design generative #AI models capable of creating new molecules, much in the same way that other generative models generate text or images? In a new paper in
@natmachintell.nature.com, we introduce #CoCoGraph, which does just that
Online: dx.doi.org/10.1038/s422...
PDF: rdcu.be/fgNIA
Can we design generative #AI models capable of creating new molecules, much in the same way that other generative models generate text or images? In a new paper in
@natmachintell.nature.com, we introduce #CoCoGraph, which does just that
Online: dx.doi.org/10.1038/s422...
PDF: rdcu.be/fgNIA
ChemEmbed: a deep learning framework for metabolite identification using enhanced MS/MS data and multidimensional molecular embeddings url: academic.oup.com/bib/article/...
The collaborative constrained graph diffusion model CoCoGraph generates novel molecules that are guaranteed to be valid and more realistic than state-of-the-art outputs, while achieving faster perform...
www.nature.com
Finding and developing new molecules is one of the great research endeavors of modern chemistry. From the development of new drugs to the creation of more sustainable materials, everything depends on finding new combinations of atoms with useful properties.
The collaborative constrained graph diffusion model CoCoGraph generates novel molecules that are guaranteed to be valid and more realistic than state-of-the-art outputs, while achieving faster performance with up to an order of magnitude fewer parameters.
dx.doi.org
The collaborative constrained graph diffusion model CoCoGraph generates novel molecules that are guaranteed to be valid and more realistic than state-of-the-art outputs, while achieving faster performance with up to an order of magnitude fewer parameters.
Abstract. Machine learning offers a promising path to annotating the large number of unidentified MS/MS spectra in metabolomics, addressing the limited cov
