Equi-mRNA, DeepBio's codon-level equivariant language model for mRNA, has been peer-reviewed and accepted at NeurIPS 2025, the 39th Conference on Neural Information Processing Systems.
The model encodes synonymous-codon symmetry directly into its architecture, as cyclic subgroups of SO(2), rather than leaving a language model to infer that structure from data. It's the representation underneath expression prediction, stability assessment, generation, and therapeutics design in the co-scientist.
For the full architecture, benchmarks against named baselines, and biological validation, see the full write-up or the model card.
