aiqbench analysis
This research introduces Pretrained Embedding Distance (PED), a scalable similarity metric derived directly from pretrained molecular models that eliminates the need for task-specific training or expensive data curation. By effectively capturing rich structural information, PED offers a computationally efficient alternative to traditional fingerprint-based methods for virtual screening and guiding molecular generation in AI-aided drug discovery. The approach addresses key bottlenecks in generality and cost, making it a practical advancement for modern computational chemistry workflows.
Advancing Ligand-based Virtual Screening and Molecular Generation with Pretrained Molecular Embedding Distance
arXiv:2604.24474v2 Announce Type: replace Abstract: Molecular similarity plays a central role in ligand-based drug discovery, such as virtual screening, analog searching, and goal-directed molecular generation. However, traditional similarity measures, ranging from fingerprint-based Tanimoto coefficients to 3D shape overlays, are often computationally expensive at scale or rely on hand-crafted molecular descriptors. Meanwhile, many deep learning approaches to similarity-aware design still depend on similarity-specific supervision or costly data curation, limiting their generality across targets. In this work, we propose pretrained embedding distance (PED) as an effective alternative, computed directly from pretrained molecular models without task-specific training. Experimental results show that PED exhibits distinct correlations with traditional similarity metrics, and performs effectively in both ranking molecules for virtual screening and guiding molecular generation via reward design. These findings suggest that pretrained molecular embeddings capture rich structural information and can serve as a promising and scalable similarity measurement for modern AI-aided drug discovery.
Abstract:Molecular similarity plays a central role in ligand-based drug discovery, such as virtual screening, analog searching, and goal-directed molecular generation. However, traditional similarity measures, ranging from fingerprint-based Tanimoto coefficients to 3D shape overlays, are often computationally expensive at scale or rely on hand-crafted molecular descriptors. Meanwhile, many deep learning approaches to similarity-aware design still depend on similarity-specific supervision or costly data curation, limiting their generality across targets. In this work, we propose pretrained embedding distance (PED) as an effective alternative, computed directly from pretrained molecular models without task-specific training. Experimental results show that PED exhibits distinct correlations with traditional similarity metrics, and performs effectively in both ranking molecules for virtual screening and guiding molecular generation via reward design. These findings suggest that pretrained molecular embeddings capture rich structural information and can serve as a promising and scalable similarity measurement for modern AI-aided drug discovery.
Accepted by ICML 2026 AI4Science (this https URL). Code and data are available
Machine Learning (cs.LG)
arXiv:2604.24474 [cs.LG]
(or arXiv:2604.24474v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2604.24474 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 249 words