Revealing the grammar of small RNA secretion using interpretable machine learning

Aug 3, 2026

Upload by Aurora

Small non-coding RNAs are secreted via mechanisms such as exosomal sorting, small extracellular vesicles, and lipoprotein complexes, yet the cis-regulatory logic controlling their extracellular sorting and release remains insufficiently defined. The study introduces ExoGRU, an interpretable machine-learning model that predicts small RNA secretion probabilities directly from primary RNA sequences, supported by experimental validation. It further introduces exoCLIP to uncover RNA–RBP interactions in the extracellular space and uses ExoGRU-guided design to create synthetic RNAs with high secretion ability, enabling cis and trans factor discovery for therapeutic and synthetic biology.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages21

Reading Time53 min

AccessFree to download

Summary

ExoGRU model and prediction framework

ExoCLIP and extracellular RNA–RBP interactions

Experimental validation and synthetic RNA design

Introduction

Background on extracellular small non-coding RNAs

Knowledge gaps in sorting and secretion mechanisms

Development of ExoGRU for cis-regulatory grammar discovery

UCSF UC San Francisco Previously Published Works Title Revealing the grammar of small RNA secretion using interpretable machine learning. Permalink [https://escholarship.org/uc/item/3zm0f0b4](https://escholarship.org/uc/item/3zm0f0b4) Journal Cell Genomics, 4(4) Authors Zirak, Bahar Naghipourfar, Mohsen Saberi, Ali et al. Publication Date 2024-04-10 DOI 10.1016/j.xgen.2024.100522 Peer reviewed [eScholarship.org](eScholarship.org) Powered by the California Digital Library University of California Article Revealing the grammar of small RNA secretion using interpretable machine learning Authors Bahar Zirak, Mohsen Naghipourfar, Ali Saberi, ..., Albertas Navickas, Ali Shariﬁ-Zarchi, Hani Goodarzi Correspondence [albertas.navickas@curie.fr](albertas.navickas@curie.fr) (A.N.), ashariﬁ[z@gmail.com](z@gmail.com) (A.S.-Z.), [hani.goodarzi@ucsf.edu](hani.goodarzi@ucsf.edu) (H.G.) In brief Zirak et al. introduce ExoGRU, a machine learning model predicting small RNA secretion from primary RNA sequences. Additionally, the study introduces exoCLIP, unveiling RNA and RNA-binding protein interactions in the extracellular space. These insights hold promise for therapeutic and synthetic biology applications. d ExoGRU utilizes primary sequences of small RNAs to accurately predict their secretion d ExoGRU identiﬁes both known and novel RBPs governing small RNA secretion d Using ExoGRU, we generated synthetic small RNAs with high secretion ability Zirak et al., 2024, Cell Genomics 4, 100522 April 10, 2024 ª 2024 The Authors. [https://doi.org/10.1016/j.xgen.2024.100522](https://doi.org/10.1016/j.xgen.2024.100522) ll ll OPEN ACCESS Article Revealing the grammar of small RNA secretion using interpretable machine learning Bahar Zirak,1,2,3,4,15 Mohsen Naghipourfar,5,15 Ali Saberi,6,7,15 Delaram Pouyabahar,8,9,15 Amirhossein Zarezadeh,10,11,15 Lixi Luo,1,2,3,4,12,15 Lisa Fish,1,2,3,4 Doowon Huh,13 Albertas Navickas,1,2,3,4,14,* Ali Shariﬁ-Zarchi,5,* and Hani Goodarzi1,2,3,4,16,* 1Department of Biochemistry & Biophysics, University of California, San Francisco, San Francisco, CA, USA 2Department of Urology, University of California, San Francisco, San Francisco, CA, USA 3Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, San Francisco, CA, USA 4Bakar Computational Health Sciences Institute, University of California, San Francisco, San Francisco, CA, US 5Department of Computer Engineering, Sharif University of Technology, Tehran, Iran 6Department of Electrical and Computer Engineering, McGill University, Montreal, QC H3A 0E9, Canada 7McGill Genome Centre, Victor Phillip Dahdaleh Institute of Genomic Medicine, 740 Dr Penﬁeld Avenue, Montreal, QC H3A 0G1, Canada 8Department of Molecular Genetics, University of Toronto, Toronto, ON, Canada 9The Donnelly Centre, University of Toronto, Toronto, ON, Canada 10Department of Stem Cells and Developmental Biology, Cell Science Research Center, Royan Institute for Stem Cell Biology and Technology, ACECR, Tehran, Iran 11Department of Developmental Biology, School of Basic Sciences and Advanced Technologies in Biology, University of Science and Culture, Tehran, Iran 12Department of Surgical Oncology, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou, China 13Laboratory of Systems Cancer Biology, The Rockefeller University, New York, NY, USA 14Institut Curie, CNRS UMR3348, INSERM U1278, Orsay, France 15These authors contributed equally 16Lead contact *Correspondence: [albertas.navickas@curie.fr](albertas.navickas@curie.fr) (A.N.), ashariﬁ[z@gmail.com](z@gmail.com) (A.S.-Z.), [hani.goodarzi@ucsf.edu](hani.goodarzi@ucsf.edu) (H.G.) [https://doi.org/10.1016/j.xgen.2024.100522](https://doi.org/10.1016/j.xgen.2024.100522) SUMMARY Small non-coding RNAs can be secreted through a variety of mechanisms, including exosomal sorting, in small extracellula

FAQs

How was the model validated experimentally?

The study validates ExoGRU performance using ExoGRU-guided mutagenesis and synthetic RNA sequence analysis. It also uses the model to identify cis and trans factors, including known and novel RBPs.

Aug 3, 2026

Upload by Aurora