Domain-Adapted Molecular Language Models for Efficient Search of Make-on-Demand Libraries

arXiv cs.LG Papers

Summary

This paper benchmarks pretrained molecular language models for virtual screening and demonstrates that explicit domain adaptation improves their performance and sample efficiency across drug discovery, materials, and catalysis libraries.

arXiv:2608.17567v1 Announce Type: new Abstract: Pretrained molecular language models are increasingly used as molecular encoders for learning structure-property relationships. However, their practical suitability for molecular discovery within and beyond their pretraining domain remains unclear. Herein, we systematically benchmark four molecular language models across six virtual molecular libraries spanning drug discovery, organic materials, and catalysis. Native molecular language model embeddings show substantial variation in discovery performance across libraries, whereas molecular fingerprints provide a consistently strong and robust baseline. Consistent with a potential domain-representation mismatch, we show that explicit domain adaptation substantially improves representation performance. Fine-tuning molecular language model encoders on structures from the target virtual library consistently improves sample efficiency, with several adapted encoders emerging as the top-performing representations across the benchmark tasks. These results show that molecular representation quality depends strongly on the target domain and that explicit adaptation can improve the practical utility of molecular foundation models. More broadly, our findings establish domain-adapted molecular representations as a promising strategy for sample-efficient adaptive decision making in virtual screening and self-driving laboratories.
Original Article
View Cached Full Text

Cached at: 08/19/26, 10:29 AM

# Domain-Adapted Molecular Language Models for Efficient Search of Make-on-Demand Libraries
Source: [https://arxiv.org/abs/2608.17567](https://arxiv.org/abs/2608.17567)
[View PDF](https://arxiv.org/pdf/2608.17567)

> Abstract:Pretrained molecular language models are increasingly used as molecular encoders for learning structure\-property relationships\. However, their practical suitability for molecular discovery within and beyond their pretraining domain remains unclear\. Herein, we systematically benchmark four molecular language models across six virtual molecular libraries spanning drug discovery, organic materials, and catalysis\. Native molecular language model embeddings show substantial variation in discovery performance across libraries, whereas molecular fingerprints provide a consistently strong and robust baseline\. Consistent with a potential domain\-representation mismatch, we show that explicit domain adaptation substantially improves representation performance\. Fine\-tuning molecular language model encoders on structures from the target virtual library consistently improves sample efficiency, with several adapted encoders emerging as the top\-performing representations across the benchmark tasks\. These results show that molecular representation quality depends strongly on the target domain and that explicit adaptation can improve the practical utility of molecular foundation models\. More broadly, our findings establish domain\-adapted molecular representations as a promising strategy for sample\-efficient adaptive decision making in virtual screening and self\-driving laboratories\.

## Submission history

From: Felix Strieth\-Kalthoff \[[view email](https://arxiv.org/show-email/45f349fc/2608.17567)\] **\[v1\]**Tue, 18 Aug 2026 09:27:56 UTC \(15,438 KB\)

Similar Articles

Large Discovery Models: Empirically-grounded Model-Based Open-Ended Search

Hugging Face Daily Papers

The paper introduces Large Discovery Model (LDM), a recurrent architecture that couples generative models with Bayesian non-parametric surrogates to guide uncertainty-aware search in scientific domains like molecules and proteins, achieving significant performance gains over existing methods.