Cure8 research brief
Cure8 research brief
A reusable transcriptomic atlas could let researchers compare new IBD biopsy datasets to a common reference, improving cross-study comparisons and helping separate inflammation from persistent disease-state biology that may remain during remission.
Researchers developing molecular biomarkers, clinicians and translational teams working on IBD tissue-based studies, and trialists using biopsy transcriptomics (including those studying therapies like olamkicept).
This is a preprint reporting a computational reference (IBDex) for reusing historical bulk biopsy RNA-seq data. The authors trained a conditional variational autoencoder on 3,168 biopsies and froze the model so new datasets can be projected into the same molecular space.
In held-out data the atlas reduced cohort-specific signatures while preserving anatomy, diagnosis, and inflammation signals; it also separated inflammation from underlying disease-state biology. The atlas was applied without retraining to independent cohorts and to samples from an olamkicept trial, tracking responder trajectories.
The authors report that deep clinical remission made inflammation look more like controls molecularly, but a persistent disease-state signature remained displaced from controls.
Preprint on medRxiv; results are computational and need peer review and independent validation before clinical application. The atlas is based on bulk tissue RNA-seq, which mixes signals from multiple cell types (single-cell methods were used for comparison but the atlas itself operates on bulk data).
Review the original publication for the complete reporting, methods, and context.
This Cure8 brief is based on source text from the linked article. Cure8 is informational only and is not a substitute for professional medical advice, diagnosis, or treatment.