Work Package 1

Strategy

The main aim of this WP is to generate, improve and refine the algorithms capable of predicting response to biological treatment based on epigenetic DNA methylation markers that we found to discriminate responders from non-responders to treatment with 3 different mainstream biologicals in CD. Further, we also test the validity of this approach in Rheumatoid Arthritis (RA) and Psoriasis (PsO). This work is built on an earlier study in 240 patients (performed together with researchers from Radcliff Hospital at Oxford University) in it was found that epigenetic biomarkers can indeed be used to guide personalized medication in these chronic immune diseases.

Output

Calibrated and validated models for each biological used for CD, RA and PsO.

Innovative elements

• An algorithm based on Machine Learning is created by our small and medium enterprises partner Horaizon and is unique and state-of-the-art in its kind, reflected by multiple high impact publications in the field
• Algorithms not only predict CD therapy success (to be validated in WP3) but will also be developed for RA/PsO -in WP4- as auto-immune diseases with a similar unpredictable biological treatment efficacy.

Partners

A sound combination of geneticists, bioinformaticians (Amsterdam University Medical Centres – location AMC and University of Oxford – Nuffield Department of Experimental Medicine Division) and a deep-tech company (Horaizon) with a solid track record in machine learning methods applied in previous CD and RA cohorts.