INTRODUCING HE2RNAPredict RNA-seq expression from digital pathology data with deep learning

Access the world’s first pan-cancer model to predict transcriptomic data directly from a tissue sample

A new tool for biomedical researchers

HE2RNA is a cutting edge tool allowing academic researchers to:

  • Generate genomic profiles from digital pathology data – Predict expression of up to 3.000 protein-encoding genes, in 28 different tumor types
  • Spatialize the transcriptomic prediction directly on the whole slide image to identify new biomarkers
  • Foster collaboration between biomedical and data science teams: improve predicitve models by transferring genomic information learned from other datasets

HE2RNA was published in Nature Communications in August 2020

Explore HE2RNA
Use cases for diverse research settings

Gene expression levels
Predict from 3000 protein-coding genes, in 28 cancer types
New genomic pathways
Correlate RNA-Seq profiles with morphological & clinical features
Knowledge transfer
Use pre-trained transcriptomic representations on small datasets

Introducing HE2RNA in Owkin Studio

Watch our lead biologist Elodie Pronier demonstrate HE2RNA in our medical research software platform Owkin Studio.

“Understanding the relationship between genotype and phenotype is one of the biggest 21st century challenges in biology. Our research opens a new path to better connect information at the genomic, cellular and tissue levels, and this would not have been possible without recent advances in artificial intelligence.”

– Gilles Wainrib, Chief Scientific Officer and Co-Founder of Owkin