A Systematic Survey and Benchmark of Deep Learning for Molecular Property Prediction in the Foundation Model Era

TL;DR

This paper provides a comprehensive survey and benchmark of deep learning methods for molecular property prediction, emphasizing recent foundation models and proposing future research directions.

Contribution

It introduces a unified taxonomy, analyzes current benchmarks, discusses challenges, and suggests new directions like physics-aware learning and multimodal datasets.

Findings

Benchmark analyses reveal inconsistencies in data curation and evaluation.

Current standards face reproducibility issues due to heterogeneous data sources.

Proposed future directions aim to improve model trustworthiness and realism.

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 datasets and datasets 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.