Learning the language of viral evolution and escape | bioRxiv
Viral mutation that escapes from human immunity remains a major obstacle to antiviral and vaccine development. While anticipating escape could aid rational therapeutic design, the complex rules governing viral escape are challenging to model. Here, we demonstrate an unprecedented ability to predict viral escape by using machine learning algorithms originally developed to model the complexity of human natural language. Our key conceptual advance is that predicting escape requires identifying mutations that preserve viral fitness, or “grammaticality,” and also induce high antigenic change, or “semantic change.” We develop viral language models for influenza hemagglutinin, HIV Env, and SARS-CoV-2 Spike that we use to construct antigenically meaningful semantic landscapes, perform completely unsupervised prediction of escape mutants, and learn structural escape patterns from sequence alone. More profoundly, we lay a promising conceptual bridge between natural language and viral evolution. One sentence summary Neural language models of semantic change and grammaticality enable unprecedented prediction of viral escape mutations. ### Competing Interest Statement The authors have declared no competing interest.
New Results doi: https://doi.org/10.1101/2020.07.08.193946 Abstract Viral mutation that escapes from human immunity remains a major obstacle to antiviral and vaccine development. While anticipating escape could aid rational therapeutic design, the complex rules governing viral escape are challenging to model. Here, we demonstrate an unprecedented ability to predict viral escape by using machine learning algorithms originally developed to model the complexity of human natural language. Our key conceptual advance is that predicting escape requires identifying mutations that preserve viral…
saved by
related reading
- On the Biology of a Large Language Modeltransformer-circuits.pub
- Language models of protein sequences at the scale of evolution enable accurate structure predictionbiorxiv.org
- Towards the explainability of protein language modelsnature.com
- [2512.11949] Neural Chameleons: Language Models Can Learn to Hide Their Thoughts from Unseen Activation Monitorsarxiv.org
- [2005.14165] Language Models are Few-Shot Learnersarxiv.org
- Interpreting Language Model Parametersgoodfire.ai
- [2511.08579] Training Language Models to Explain Their Own Computationsarxiv.org
- [2005.11401] Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasksarxiv.org
- Sparse Autoencoders Find Highly Interpretable Features in Language Modelsarxiv.org
- [2210.13382] Emergent World Representations: Exploring a Sequence Model Trained on a Synthetic Taskarxiv.org
- Predicting Empirical AI Research Outcomes with Language Modelsarxiv.org
- A socratic dialogue over the utility of DNA language models (Part 1 of 2)owlposting.com