A Systematic Survey and Benchmark of Deep Learning for Molecular Property Prediction in the Foundation Model Era
Top Cited Papers
- 7 May 2026
- journal article
- review article
- Published by American Chemical Society (ACS) in Journal of Chemical Theory and Computation
- Vol. 22 (10) , 4866-4887
Abstract
Molecular property prediction integrates quantum chemistry, cheminformatics, and deep learning to connect molecular structure with physicochemical and biological behavior. This survey traces four complementary paradigms, including Quantum, Descriptor Machine Learning, Geometric Deep Learning, and Foundation Models, and outlines a unified taxonomy linking molecular representations, model architectures, and interdisciplinary applications. Benchmark analyses integrate evidence from both widely used data sets and data sets reflecting industry perspectives, encompassing quantum, physicochemical, physiological, and biophysical domains. The survey examines current standards in data curation, splitting strategies, and evaluation protocols, highlighting challenges including inconsistent stereochemistry, heterogeneous assay sources, and reproducibility limitations under random or poorly defined splits. These observations motivate the modernization of benchmark design toward more transparent, time- and scaffold-aware methodologies. We further propose three forward-looking directions: (i) physics-aware learning embedding quantum consistency, (ii) uncertainty-calibrated foundation models for trustworthy inference, and (iii) realistic multimodal benchmark ecosystems integrating computational and experimental data. Repository: https://github.com/Zongru-Li/Survey-and-Benchmarks-of-DL-for-Molecular-Property-Prediction-in-the-Foundation-Model-Era.Keywords
All Related Versions
Funding Information
- Shenzhen-Hong Kong-Macao Science and Technology Plan Project (SGDX20210823103537030)
This publication has 103 references indexed in Scilit:
- Advancements in Molecular Property Prediction: A Survey of Single and Multimodal ApproachesArchives of Computational Methods in Engineering, 2025
- Uni‐Electrolyte: An Artificial Intelligence Platform for Designing Electrolyte Molecules for Rechargeable BatteriesAngewandte Chemie International Edition, 2025
- Scaffold Splits Overestimate Virtual Screening Performance2024
- MOL-MOE: Learning Drug Molecular Characterization Based on Mixture of Expert Mechanism2024
- Pre-training molecular representation model with spatial geometry for property predictionComputational Biology and Chemistry, 2024
- Holo-Mol: An explainable hybrid deep learning framework for predicting reactivity of hydroxyl radical to water contaminants based on holographic fused molecular representationsChemical Engineering Journal, 2024
- Molecular geometric deep learningCell Reports Methods, 2023
- Directed message passing based on attention for prediction of molecular propertiesComputational Materials Science, 2023
- AB ‐Bind: Antibody binding mutational database for computational affinity predictionsProtein Science, 2015
- The performance of density functional and wavefunction-based methods for 2D and 3D structures of Au10Journal of Computational Chemistry, 2013