.png){kind=icon}
.png){kind=icon}
Return
RNA regulation mechanisms study using physics-informed machine learning
.png){kind=icon}
OA
AI
DOI：10.1038/s42005-025-02452-z.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Understanding the dynamic regulation mechanisms of RNA is crucial for unlocking the secrets of biological processes and diseases. Existing computational methods primarily focus on static structures and give insufficient attention to RNA. Here, we develop a pipeline based on physics-informed machine learning to uncover regulatory mechanisms and manage the function of RNA complexes. We applied this pipeline to investigate the latent regulatory networks with key regions in response to regulation and detect binding sites underlying the P-TEFb/Tat/TAR system involved in HIV-1 transcriptional activation and the aaRS/tRNA system essential for genetic translation. Additionally, we search for potential small-molecule inhibitors at these sites. By integrating neural relational inference with network models, the pipeline demonstrates high performance, as evidenced by experimental studies, outperforming existing state-of-the-art tools. This pipeline serves as a paradigm for studying regulatory mechanisms in macromolecular RNA complexes, showing the insights provided by physics-informed machine learning in revealing dynamic regulatory processes. Understanding RNA regulation mechanisms is crucial for developing targeted therapies for diseases and elucidating how cells control gene expression. This study presents a physics-informed machine learning framework to investigate dynamic regulatory mechanisms in RNA complexes, which when applied to the P-TEFb/Tat/TAR system involved in HIV-1 transcriptional activation and the aaRS/tRNA system essential for genetic translation, outperforms existing methods for uncovering latent regulatory networks.
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
5.8
Papers:
2.8K
Citations:
9.2K