The field of molecular generation and interaction prediction is rapidly advancing, with a focus on developing innovative methods for predicting chemical reactions, generating synthesizable molecules, and modeling complex biological interactions. Recent studies have introduced novel graph-based approaches, such as SynBridge and MolecBioNet, which demonstrate state-of-the-art performance in bidirectional reaction prediction and drug-drug interaction prediction. Additionally, the development of conditional generative models, like SAFE-T and ToDi, has shown promise in prioritizing and designing molecules with specific biological objectives. Noteworthy papers in this area include SynBridge, which proposes a bidirectional flow-based generative model for reaction prediction, and ToDi, which introduces a generative framework for hit-like molecular generation conditioned on omics expressions and molecular textual descriptions. Overall, these advancements have the potential to accelerate drug discovery and improve our understanding of complex biological systems.
Advances in Molecular Generation and Interaction Prediction
Sources
SynBridge: Bridging Reaction States via Discrete Flow for Bidirectional Reaction Prediction
ToxBench: A Binding Affinity Prediction Benchmark with AB-FEP-Calculated Labels for Human Estrogen Receptor Alpha
Towards Interpretable Drug-Drug Interaction Prediction: A Graph-Based Approach with Molecular and Network-Level Explanations
La-Proteina: Atomistic Protein Generation via Partially Latent Flow Matching
Do we need equivariant models for molecule generation?
Efficient Molecular Conformer Generation with SO(3)-Averaged Flow Matching and Reflow
TextOmics-Guided Diffusion for Hit-like Molecular Generation
Conditional Chemical Language Models are Versatile Tools in Drug Discovery
Anticipating the Selectivity of Cyclization Reaction Pathways with Neural Network Potentials
Enhancing Safe and Controllable Protein Generation via Knowledge Preference Optimization
Unraveling the Biomarker Prospects of High-Altitude Diseases: Insights from Biomolecular Event Network Constructed using Text Mining
A Graph-in-Graph Learning Framework for Drug-Target Interaction Prediction
Torsional-GFN: a conditional conformation generator for small molecules
SynCoGen: Synthesizable 3D Molecule Generation via Joint Reaction and Coordinate Modeling
Assay2Mol: large language model-based drug design using BioAssay context