Generalizable mutation-effect prediction across adaptive immune recognition via unified multimodal framework

2026-05-27 · Source: Nature Machine Intelligence · Field: Science & Research — Life Sciences & Biology, Mathematics & Computational Sciences, Health & Medical Research · Depth: Expert, long

Summary

UniAIR is a novel, modular, multimodal framework designed for accurate and generalizable prediction of mutation effects across diverse immune recognition scenarios. This framework addresses the challenge of predicting how mutations alter antigen-lymphocyte interactions, a critical aspect of adaptive immunity. UniAIR integrates a standardized data pipeline, an interface-centric sequence-structure fusion transformer that combines evolutionary information with geometric representations, and extensions for multiexpert consensus and adaptation to predicted structure inputs. Comprehensive evaluation through large-scale benchmarking and independent tests demonstrated UniAIR's superior performance, achieving results comparable to the best existing methods across tasks like antibody maturation, antigen escape, and TCR-pHLA optimization, even when experimental structures were unavailable. It successfully performed multiround peptide optimization of a TCR-pHLA complex under sparse feedback and identified key functional mutations in incomplete antibody-antigen structures, establishing a unified computational foundation for immunotherapeutic design.

Key takeaway

For AI Scientists and Research Scientists developing immunotherapies, UniAIR offers a robust computational foundation for predicting mutation effects. You should consider integrating this multimodal framework to accelerate the design of novel therapeutics by accurately mapping mutation landscapes. This approach can optimize high-affinity peptide mutants and identify critical functional mutations, even with incomplete structural data, streamlining your drug discovery pipeline.

Key insights

UniAIR unifies multimodal data and structural information to predict immune mutation effects with high generalizability.

Principles

• Integrate evolutionary and geometric data for robust predictions.
• Multimodal frameworks enhance generalization across immune tasks.
• Adapt to predicted structures for broader applicability.

Method

UniAIR employs a standardized data pipeline, an interface-centric sequence-structure fusion transformer, and multiexpert consensus for mutation-effect prediction. It also adapts to predicted structure inputs.

In practice

• Optimize peptide mutants for high-affinity TCR-pHLA complexes.
• Identify functional mutations in incomplete antibody-antigen structures.
• Map mutation landscapes for immunotherapeutic design.

Topics

• Adaptive Immunity
• Mutation Effect Prediction
• Multimodal AI
• Protein Structure Prediction
• Immunotherapeutic Design
• TCR-pHLA Optimization

Code references

• hanrthu/UniAIR

Best for: AI Scientist, Research Scientist, Machine Learning Engineer

Related on AIssential

• Bridging the phenotype-target gap for molecular generation via multi-objective reinforcement learning — ExMolRL combines phenotype-guided generation with target-aware reinforcement learning for de novo drug discovery.
• AgForce Enables Antigen-conditioned Generative Antibody Design — Existing antibody design models fail to condition on antigen, suffering from antigen blindness and vocabulary collapse due to cross-entropy loss.
• GigaTIME: Scaling tumor microenvironment modeling using virtual population generated by multimodal AI — GigaTIME uses multimodal AI to convert common H&E slides into virtual mIF images, enabling population-scale precision immuno-oncology research.
• UniRG: Scaling medical imaging report generation with multimodal reinforcement learning — Reinforcement learning with clinically meaningful rewards significantly improves medical vision-language model reliability and generality.
• An Evolutionary Approach for Designing Stable and Highly Expressible Low-Immunogenicity Therapeutic mRNA Sequences — The BERT-GA framework optimizes therapeutic mRNA sequences for balanced translation efficiency, structural stability, and low immunogenicity.
• AlphaGenome author roundtable — AlphaGenome unifies DNA sequence-to-function prediction, deciphering genetic variant impact across the entire human genome.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Nature Machine Intelligence.