TORONTO, May 19, 2025, Deep Genomics, the AI-first TechBio organization pioneering a new path for genomic R&D, today announced the latest addition to its foundation model platform. The new deep learning model, REPRESS, accurately predicts microRNA (miRNA) binding and mRNA degradation directly from RNA sequences. The model marks a major advancement in post-transcriptional biology and therapeutic design.
mRNA stability is crucial for maintaining normal protein levels and boosting protein levels, both of which offer therapeutic benefits. REPRESS enables the fast and accurate analyses of patient mutations and therapeutic approaches that act through mRNA stability.
Trained on millions of miRNA-mRNA interaction and degradation sites from human and mouse tissues, REPRESS (Regulatory Element PRediction of post-transcriptional Events using Sequence Signals) reveals regulatory mechanisms previously hidden, even from state-of-the-art technologies. By learning to predict miRNA binding and mRNA degradation directly from sequence, it sheds light on a critical and underexplored layer of gene regulation.
Lead scientist Bhargav Kanuparthi of Deep Genomics said: ”REPRESS is the first foundation model of its kind and predicts therapeutically relevant mechanisms driving mRNA stability features like miRNA binding and mRNA degradation with substantially more accuracy than other methods. Now that it’s incorporated into our Foundation Model Platform, we are excited to see how it can benefit target identification and the design of ASO, siRNA, and mRNA therapeutics.”
REPRESS outperforms seven leading methods across seven distinct benchmarks, including variant effect prediction, generalization to reporter assay data, and modeling both canonical and non-canonical miRNA repression.
“REPRESS is the latest example of the innovation that happens when AI and biology are deeply integrated,” said Dr. Brendan Frey, Founder and CIO of Deep Genomics. “By offering a new lens into a previously obscure layer of gene regulation, REPRESS brings us closer to designing more precise and effective therapeutics. I’m incredibly proud of the Deep Genomics team whose talent and dedication made this breakthrough possible.”
REPRESS is a key component of Deep Genomics’ cohesive, multi-model Foundation Model Platform, which accelerates therapeutic discovery across a broad spectrum of RNA modalities. It is complementary with Deep Genomics’ other proprietary models for RNA processing and RNA editing to form a unified system for decoding complex gene regulation. The models collaborate, enabling more comprehensive target identification and therapeutic design.
REPRESS is being made freely available for non-commercial use via Deep Genomics’ GitHub repository. For more information and to access the model, visit https://github.com/deepgenomics/repress
About Deep Genomics
Deep Genomics is an AI-first TechBio organization pioneering a new path for genomic R&D through its multidisciplinary team of AI and biology scientists and its Foundation Model Platform. Leveraging cutting-edge AI technologies and biological insights, Deep Genomics is at the forefront of revolutionizing drug discovery and development. Through its innovative approach and collaborative partnerships, Deep Genomics aims to accelerate the delivery of life-saving therapeutics to patients worldwide. Deep Genomics is located in Toronto, Ontario, and Cambridge, Massachusetts. For more information, visit www.deepgenomics.com and follow us on LinkedIn.
Media Contact:
press@deepgenomics.com
Follow us on LinkedIn