State-specific protein–ligand complex structure prediction with a multiscale deep generative model

By Zhuoran Qiao, Weili Nie, Arash Vahdat, Thomas F. Miller, Anima Anandkumar. 2022

173 citations recorded by the source corpus.

The binding complexes formed by proteins and small molecule ligands are ubiquitous and critical to life. Despite recent advancements in protein structure prediction, existing algorithms are so far unable to systematically predict the binding ligand structures along with their regulatory effects on protein folding. To address this discrepancy, we present NeuralPLexer, a computational approach that can directly predict protein–ligand complex structures solely using protein sequence and ligand molecular graph inputs. NeuralPLexer adopts a deep generative model to sample the three-dimensional structures of the binding complex and their conformational changes at an atomistic resolution. The model is based on a diffusion process that incorporates essential biophysical constraints and a multiscale geometric deep learning system to iteratively sample residue-level contact maps and all heavy-atom coordinates in a hierarchical manner. NeuralPLexer achieves state-of-the-art performance compared with all existing methods on benchmarks for both protein–ligand blind docking and flexible binding-site structure recovery. Moreover, owing to its specificity in sampling both ligand-free-state and ligand-bound-state ensembles, NeuralPLexer consistently outperforms AlphaFold2 in terms of global protein structure accuracy on both representative structure pairs with large conformational changes and recently determined ligand-binding proteins. NeuralPLexer predictions align with structure determination experiments for important targets in enzyme engineering and drug discovery, suggesting its potential for accelerating the design of functional proteins and small molecules at the proteome scale. Great advances in protein structure prediction have been made with recent deep learning-based methods, but proteins interact with their environment and can change shape drastically when binding to ligand molecules. To predict the 3D structure of these combined protein–ligand complexes, Qiao et al. developed a generative diffusion model with biophysical constraints and geometric deep learning.

Subfields: Molecular Biology, Generative Models, Computational Chemistry & Drug Discovery

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

1. Pushmeet Kohli
2. Alexander Pritzel
3. Eirini Arvaniti
4. Michael Figurnov
5. Ellen D. Zhong

In Generative Models

1. Anna Potapenko
2. Pushmeet Kohli
3. Alexander Pritzel
4. Demis Hassabis
5. Michael Figurnov

Citations in the corpus

• Accurate structure prediction of biomolecular interactions with AlphaFold 3 (2024, methodology): “A wide range of predictors for various specific interaction types has been developed 16–28 , as well as one generalist method developed concurrently with the present work 29 , but the accuracy of such deep-learning attempts has been mixed and often below that of physics-inspired methods 30,31 .”
• Scalable emulation of protein equilibrium ensembles with generative deep learning (2025, unknown): “Such models [13–15], as well as perturbation-based derivatives of AlphaFold [16, 17] have also been shown to be capable of generating distinct protein structures and can be combined with MD simulation to alleviate the sampling problem [18].”
• Artificial intelligence in drug development (2025, unknown): “AI-based receptor–ligand docking models can predict ligand spatial transformations, directly generate complex atomic coordinates using algorithms like equivariant neural networks 44–46 and learn the probability density distribution of receptor– ligand distances to generate binding poses 47–49 .”