AROMA: Augmented Reasoning Over a Multimodal Architecture for Virtual Cell Genetic Perturbation Modeling

ACL ARR 2026 January Submission2274 Authors

Keywords: Virtual cell perturbation prediction, Knowledge-graph grounded multimodal modeling, Retrieval-augmented reasoning

Abstract: Virtual cell modeling predicts molecular state changes under genetic perturbations in silico, and is essential for biological mechanism studies. However, existing approaches suffer from unconstrained reasoning, uninterpretable predictions, and retrieval signals that are weakly aligned with regulatory topology. To address these limitations, we propose AROMA, an Augmented Reasoning Over a Multimodal Architecture for virtual cell genetic perturbation modeling. AROMA integrates textual evidence, graph-topology information, and protein sequence features to model perturbation-target dependencies, and is trained with a two-stage optimization strategy to yield predictions that are both accurate and interpretable. We also construct two knowledge graphs and a perturbation reasoning dataset, PerturbReason, containing more than 498k samples, as reusable resources for the virtual cell domain. Experiments show that AROMA outperforms existing methods across multiple cell lines, and remains robust under zero-shot evaluation on an unseen cell line, as well as in knowledge-sparse, long-tail scenarios. Overall, AROMA demonstrates that combining knowledge-driven multimodal modeling with evidence retrieval provides a promising pathway toward more reliable and interpretable virtual cell perturbation prediction. Model weights is available at an anonymous HuggingFace repository: https://huggingface.co/AnonymousModel/AROMA.

Paper Type: Long

Research Area: Clinical and Biomedical Applications

Research Area Keywords: Language Modeling, Information Retrieval and Text Mining, NLP Applications, Dialogue and Interactive, Question Answering Systems

Contribution Types: Model analysis & interpretability, NLP engineering experiment, Publicly available software and/or pre-trained models, Data resources

Languages Studied: English

Submission Number: 2274