PARNET: A CLIP-seq-based foundation model for RNA sequence representation

anshulkundaje · x · 2026-09-17

A multi-institution team from Helmholtz Munich's Computational Health Center, King's College London, the Francis Crick Institute, TU Munich and Slovenia's National Institute of Chemistry (including Gagneur, Ule and Marsico as senior authors) released the bioRxiv preprint PARNET, a foundation model for RNA sequence representation learning trained on CLIP-seq data, aiming to learn general-purpose RNA sequence embeddings for downstream genomics studies.

Original post →

More from Research

Research channel →