RNASeek establishes a unified pretrain–predict–optimize framework that connects learned RNA function to controllable de novo sequence design and provides a general strategy for engineering regulatory RNAs with desired properties.
Abstract
RNA plays central roles in regulating information flow and provides a versatile substrate for engineering biological functions. While large language models (LLMs) have transformed natural language processing and protein design, a general framework connecting RNA foundation models to functional sequence design remains limited. Here, we present RNASeek, a 1.6-billion-parameter generative foundation model built on a DeepSeek architecture and trained on a cross-phyla transcriptomic corpus for RNA sequence representation and generation. Natural-language tokens enable flexible conditional prediction and sequence design using a unified backbone. RNASeek captures species-specific transcript features and intron–exon boundaries in a zero-shot setting. We then fine-tune RNASeek to predict ribozyme self-cleavage activity and viral mRNA stability, revealing interpretable sequence features associated with function, including ribozyme loop flexibility and stem stability, as well as AU-rich motifs associated with mRNA stability. We use these functional predictors as reward models and apply Group Relative Policy Optimization (GRPO) to update RNASeek’s generation policy toward sequences with desired properties. GRPO-guided generation produces faster-cleaving ribozymes and stability-enhancing 3′ UTRs while satisfying user-specified IUPAC constraints. Experimentally validated RNASeek-generated ribozymes achieve wild-type levels of activity, while RNASeek-generated 3′ UTR sequences exceed the performance of the training data and benchmarked AI-generated 3′ UTRs. Together, RNASeek establishes a unified pretrain–predict–optimize framework that connects learned RNA function to controllable de novo sequence design and provides a general strategy for engineering regulatory RNAs with desired properties.
RNA foundation models offer a promising route to decoding the relationships between RNA sequence, structure and function, yet how model architecture shapes the acquisition and organization of biological information remains poorly understood. Here we introduce Layer-Interleaved Mamba-Bidirectional Encoder for RNA (LIMBE...
Ziyao Wang, Yun-Rui Qiu, Ye-Xiao He et al.· bioRxiv· 0 citations
RNA engineering is bottlenecked by the tight coupling among sequence, structure, and function, leaving the field reliant on heuristics or screening. We introduce RNAverse, a multimodal generative model that learns a conditionable design grammar from RNA sequences, experimental structural probing profiles, and functiona...
Protein-coding sequences are read through the triplet genetic code, whereas the organization of molecular information within non-coding RNA sequences remains elusive. Here, we identify a non-coding RNA grammar defined by multi-scale motif usage and 5’-to-3’ positional gradients, and develop RiboGRAM to learn this organ...
Hui Wang, Zhao-Ming Chen, He Huang et al.· bioRxiv· 0 citations
This tutorial and literature survey provide a comprehensive overview of machine learning methodologies applied to RRI prediction, focusing on classical models such as Naive Bayes and Bayesian networks, alongside cutting-edge deep learning techniques exemplified by Bidirectional Encoder Representations from Transformers...
S. Parthiban, Swetha Varadarajan· Methods in molecular biology· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.