Multimodal Alignment and Preference Optimization for Zero-Shot Conditional RNA Generation
Abstract
The design of RNA molecules that interact with specific proteins is a critical challenge in experimental and computational biology. Despite recent progress in natural language modeling and deep learning-based protein design, there remains significant room to improve the frequency of successful interactions and the authenticity of generated sequences for functional applications. In this work, we frame conditional RNA sequence generation as a multi-stage alignment problem, introducing Moirain: a suite of models optimized via multimodal supervised fine-tuning (SFT) and Direct Preference Optimization (DPO). Our approach begins with large-scale pretraining on diverse RNA corpora to capture the fundamental grammars of sequence plausibility. To achieve target-specific generation, we employ a multimodal SFT architecture that conditions RNA synthesis on protein structural and sequential features. Finally, we leverage DPO to refine the model using synthetic interaction data: taking advantage of DPO's unique ability to navigate non-aligned preference spaces, we improve functional fitness without collapsing the learned natural distribution. Extensive evaluation of the Moirain series (Moirain-Base, -Multi, and -DPO) demonstrates that our framework consistently produces novel, diverse, and biologically plausible RNA sequences with superior binding affinities compared to existing baselines.
Related
- SAIL: Sparse Autoencoders for Interpretable Alignment of Human Vision and Multimodal Large Language Models
- trFlow: ultrafast all-atom protein conformational ensemble generation via conditional flow matching
- Wigner–Ville-Based Conditional Autoencoder For Biosonar Generation And Joint Range And Off-Axis Estimation With A Single Hydrophone
- OmniTCR: a foundation model unifying T cell receptor recognition prediction and conditional sequence generation
- Cross-modal alignment with language tokens for unified multilingual image-to-text generation.
BibTeX
@misc{klypa2026multimodal,
title = {{Multimodal Alignment and Preference Optimization for Zero-Shot Conditional RNA Generation}},
author = {Roman Klypa and Alberto Bietti and Sergei Grudinin},
howpublished = {HAL},
year = {2026},
month = may,
eprint = {2605.23961},
archivePrefix = {arXiv},
url = {https://hal.science/hal-05734068v1},
note = {CC BY-SA 4.0},
}
RIS
TY - UNPB
TI - Multimodal Alignment and Preference Optimization for Zero-Shot Conditional RNA Generation
AU - Roman Klypa
AU - Alberto Bietti
AU - Sergei Grudinin
PY - 2026
DA - 2026/05/12
PB - HAL
UR - https://hal.science/hal-05734068v1
AB - The design of RNA molecules that interact with specific proteins is a critical challenge in experimental and computational biology. Despite recent progress in natural language modeling and deep learning-based protein design, there remains significant room to improve the frequency of successful interactions and the authenticity of generated sequences for functional applications. In this work, we frame conditional RNA sequence generation as a multi-stage alignment problem, introducing Moirain: a suite of models optimized via multimodal supervised fine-tuning (SFT) and Direct Preference Optimization (DPO). Our approach begins with large-scale pretraining on diverse RNA corpora to capture the fundamental grammars of sequence plausibility. To achieve target-specific generation, we employ a multimodal SFT architecture that conditions RNA synthesis on protein structural and sequential features. Finally, we leverage DPO to refine the model using synthetic interaction data: taking advantage of DPO's unique ability to navigate non-aligned preference spaces, we improve functional fitness without collapsing the learned natural distribution. Extensive evaluation of the Moirain series (Moirain-Base, -Multi, and -DPO) demonstrates that our framework consistently produces novel, diverse, and biologically plausible RNA sequences with superior binding affinities compared to existing baselines.
ER -