AntigenLM: Structure-Aware DNA Language Modeling for Influenza

2026/02/09 by Yue Pei, Xuebin Chi, Yu Kang · 1 voice
Biochemistry, Genetics and Molecular Biology · Computer Science · #cs.AI #cs.CE #cs.CL #q-bio.GN

paper · pdf · doi:10.48550/arxiv.2602.09067

Abstract

Language models have advanced sequence analysis, yet DNA foundation models often lag behind task-specific methods for unclear reasons. We present AntigenLM, a generative DNA language model pretrained on influenza genomes with intact, aligned functional units. This structure-aware pretraining enables AntigenLM to capture evolutionary constraints and generalize across tasks. Fine-tuned on time-series hemagglutinin (HA) and neuraminidase (NA) sequences, AntigenLM accurately forecasts future antigenic variants across regions and subtypes, including those unseen during training, outperforming phylogenetic and evolution-based models. It also achieves near-perfect subtype classification. Ablation studies show that disrupting genomic structure through fragmentation or shuffling severely degrades performance, revealing the importance of preserving functional-unit integrity in DNA language modeling. AntigenLM thus provides both a powerful framework for antigen evolution prediction and a general principle for building biologically grounded DNA foundation models.

Citations

• Integrative modeling of seasonal influenza evolution via AI-powered antigenic cartography
• Genomic Language Models: Opportunities and Challenges
• VQDNA: Unleashing the Power of Vector Quantization for Multi-Species Genomic Sequence Modeling
• HyenaDNA: Long-Range Genomic Sequence Modeling at Single Nucleotide Resolution
• Evolutionary-scale prediction of atomic-level protein structure with a language model
• Effective gene expression prediction from sequence by integrating long-range interactions
• Attention Is All You Need
• Google's Multilingual Neural Machine Translation System: Enabling Zero-Shot Translation
• Prediction, dynamics, and visualization of antigenic phenotypes of seasonal influenza viruses
• Neural Machine Translation of Rare Words with Subword Units
• Predicting evolution from the shape of genealogical trees
• Viral Evolution in the Genomic Age

Discussions

• Yue Pei, Xuebin Chi, Yu Kang: AntigenLM: Structure-Aware DNA Language Modeling for Influenza https://arxiv.org/abs/2602.09067 https://arxiv.org/pdf/2602.09067 https://arxiv.org/html/2602.09067 [bsky, 0 points, 0 comments]

Related

• Mitigating the Antigenic Data Bottleneck: Semi-supervised Learning with Protein Language Models for Influenza A Surveillance
• Unsupervised language models for disease variant prediction
• DPCIPI: A pre-trained deep learning model for predicting cross-immunity between drifted strains of Influenza A/H3N2
• Pathogen genomic surveillance and the AI revolution
• Genomic Language Models: Opportunities and Challenges