Document Type
Article
Publication Date
11-28-2025
Publication Title
Nature Communications
Abstract
The rapid evolution of DNA foundation models promises to revolutionize genomics, yet comprehensive evaluations are lacking. Here, we present a comprehensive, unbiased benchmark of five models (DNABERT-2, Nucleotide Transformer V2, HyenaDNA, Caduceus-Ph, and GROVER) across diverse genomic and genetic tasks including sequence classification, gene expression prediction, variant effect quantification, and topologically associating domain (TAD) region recognition, using zero-shot embeddings. Our analysis reveals that mean token embedding consistently and significantly improves sequence classification performance, outperforming other pooling strategies. Model performance varies among tasks and datasets; while general purpose DNA foundation models showed competitive performance in pathogenic variant identification, they were less effective in predicting gene expression and identifying putative causal QTLs compared to specialized models. Our findings offer a framework for model selection, highlighting the impact of architecture, pre-training data, and embedding strategies on performance in genomic and genetic tasks.
First Page
10780
PubMed ID
41315262
Volume
16
Issue
1
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Recommended Citation
Feng, Haonan; Wu, Lang; Zhao, Bingxin; Huff, Chad; Zhang, Jianjun; Wu, Jia; Lin, Lifeng; Wei, Peng; and Wu, Chong, "Benchmarking DNA foundation models for genomic and genetic tasks" (2025). School of Medicine Faculty Publications. 4381.
https://digitalscholar.lsuhsc.edu/som_facpubs/4381
10.1038/s41467-025-65823-8
Included in
Artificial Intelligence and Robotics Commons, Biostatistics Commons, Genomics Commons, Investigative Techniques Commons, Medical Genetics Commons