Evolutionary Informed Translation from Protein to Nucleic Acid Sequence
Abstract
The choice of nucleic acid sequence for a given protein strongly impacts function and is species-specific. Current protein-to-nucleic-acid translation models treat sequences in isolation, often over-favoring frequently used codons. To generate more natural nucleic acid sequences, we introduce PINATA (Phylogenetically Informed protein to Nucleic Acid Translation with sequence Alignment), a model that leverages sequences from related species encoded in multiple sequence alignments (MSAs) to guide prediction. Our architecture leverages evolutionary context and learns species-and protein specific context to make biologically-informed predictions. Evaluated across 61,934 gene trees and 882 species, PINATA generates sequences that mirror the naturally-occurring sequence substantially better than existing language models. PINATA leverages MSA data to improve exact native codon recovery by 17.95 percentage points over existing language models, effectively preserves natural rare-codon nuances, and more closely matches native sequence properties. PINATA maintains strong predictive performance even with limited evolutionary data. While previous methods largely treated sequences from different species in isolation, our method leverages the evolutionary relationships between nucleic acids and proteins across species to create a more biologically-informed translation. We demonstrate that leveraging rich evolutionary data through an biologically-designed method strongly aids in translating protein sequences to nucleic acid sequences across species. We will make data and code publicly available upon publication.
Related
- Sequence-dependent conformational and mechanical landscapes of double-stranded nucleic acids
- Evolutionary extensions of IF2 couple translation initiation to bacterial stress adaptation
- NuXLApp: Interactive data analysis and visualization for protein‐nucleic acid crosslinking mass spectrometry
- The MODNAP benchmark set for modeling ribose-based modified nucleic acid and nucleic acid–protein complexes: a case study with AlphaFold 3
- Sequence- and Base-Selective Modification of mRNA within and nearby Premature Termination Codon and Its Effect on the Translation in a Human Cell-Free Translation System
BibTeX
@misc{hart2026evolutionary,
title = {{Evolutionary Informed Translation from Protein to Nucleic Acid Sequence}},
author = {Anna Hart and Mohammed El-Kebir and Huimin Zhao and Heng Ji},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.22562},
eprint = {2610.22562},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.22562},
}
RIS
TY - UNPB
TI - Evolutionary Informed Translation from Protein to Nucleic Acid Sequence
AU - Anna Hart
AU - Mohammed El-Kebir
AU - Huimin Zhao
AU - Heng Ji
PY - 2026
DA - 2026/10/15
PB - arXiv
DO - 10.48550/arxiv.2610.22562
UR - https://arxiv.org/abs/2610.22562
AB - The choice of nucleic acid sequence for a given protein strongly impacts function and is species-specific. Current protein-to-nucleic-acid translation models treat sequences in isolation, often over-favoring frequently used codons. To generate more natural nucleic acid sequences, we introduce PINATA (Phylogenetically Informed protein to Nucleic Acid Translation with sequence Alignment), a model that leverages sequences from related species encoded in multiple sequence alignments (MSAs) to guide prediction. Our architecture leverages evolutionary context and learns species-and protein specific context to make biologically-informed predictions. Evaluated across 61,934 gene trees and 882 species, PINATA generates sequences that mirror the naturally-occurring sequence substantially better than existing language models. PINATA leverages MSA data to improve exact native codon recovery by 17.95 percentage points over existing language models, effectively preserves natural rare-codon nuances, and more closely matches native sequence properties. PINATA maintains strong predictive performance even with limited evolutionary data. While previous methods largely treated sequences from different species in isolation, our method leverages the evolutionary relationships between nucleic acids and proteins across species to create a more biologically-informed translation. We demonstrate that leveraging rich evolutionary data through an biologically-designed method strongly aids in translating protein sequences to nucleic acid sequences across species. We will make data and code publicly available upon publication.
ER -