Drug REpurposing with Artificial intelligence for Muscular disorderS

Project Information

DREAMS

Grant agreement ID: 101080229

DOI

10.3030/101080229

EC signature date 8 October 2023

Start date 1 November 2023

End date 31 October 2028

Funded under

• Health

Total cost

€ 6 952 110,00

EU contribution

€ 6 952 110,00

Coordinated by

CENTRE D'ETUDE DES CELLULES SOUCHES
France

Periodic Reporting for period 1 - DREAMS (Drug REpurposing with Artificial intelligence for Muscular disorderS)

Reporting period: 2023-11-01 to 2025-04-30

Summary of the context and overall objectives of the project

According to the World Health Organization, to date, fewer than 6% of rare diseases (RDs) have an approved treatment option. In light of this observation, it appears crucial to develop disruptive approaches that allow to find treatments for multiple RDs. To achieve this goal, DREAMS is developing a novel approach combining iPSCs with novel AI technologies to discover drugs which target common mechanisms of rare neuromuscular disorders (NMD). The project will focus on 5 NMDs that share common pathophysiological characteristics related to dysfunctions of autophagy and desmin disorganization.
The DREAMS project is set to deliver high-impact results across short, medium, and long-term horizons in the field of neuromuscular diseases (NMDs). By Month 60, using skeletal muscle cells derived from induced pluripotent stem cells (skMC-iPSC), DREAMS will, 1) generate up to 5 novel biomarkers of the 5 diseases, 2) perform a high throughput drug screen on all 5 diseases to identify 10–50 repurposable drugs and 3) validate 3 lead compounds in animal models. It will also deliver an innovative clinical trial design and platform, engaging over 160 professionals and 400 patients clinicians with the support of a well-renowned association AFM-Telethon. Within 3–5 years post-project, DREAMS will achieve preclinical validation of at least 2 therapies across 5 NMDs, establish 3–5 industry agreements in rare NMDs, and involve more than 500 professionals and 1600 patients. In the long term (>5 years), the project aims to deliver one clinically validated therapy, influence clinical guidelines, and mobilize >85% of professionals and patients in innovation activities, contributing to a projected €50 million investment in new therapies.
Together, these technological and conceptual innovations will allow the DREAMS consortium to accelerate the development of therapies by reducing timelines for small molecule identification (from 3-5 years to 12 months) and lowering drug development costs by 40%. Its focus on the identification of common mechanisms, so it expands the numbers of pathologies and then patients who could benefit from common molecular targets and potential treatments. Through a collaborative ecosystem of researchers, patients, and regulators, DREAMS will establish a regulatory-ready framework with the potential to transform care pathways across Europe for over 430,000 patients with rare NMDs.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

During the first reporting period (RP1), major progress was achieved. Twenty-two iPSC lines were characterized and successfully differentiated into myoblasts (Task 1.1) enabling disease modeling and biomarker identification. In the Task 1.2 we studied three muscle diseases (DMD, CNM, and EDMD2) to look at how cells recycle their components by a process called autophagy. We found similar issues in all three, including unusual levels of certain proteins (Caveolin-3, Galectin-3, and ATG16L1), confirmed in patient muscle biopsies. By reanalyzing existing genetic data (Task 1.3) we discovered biological pathways that are disrupted in these diseases. In Pompe disease, we saw problems in the cell’s waste disposal system (lysosomes), and in EDMD2 and DMD, we found structural and energy-related defects. The mitochondria, that produce the energy in the cells also didn’t work properly in all three diseases. Finally, we created a test to measure two markers (p62 and lamp1) linked to autophagy.
In WP2, the work package leader laid the technological foundations with major progress on Sapian AI technology. The Sapian.Hit model, which predicts compound-protein interactions, saw a notable performance improvement (AUPRC > 71%). Sapian.Indication was developed to map therapeutic targets to diseases across 1,167 conditions. Sapian.Target was successfully tested to identify causal targets from phenotypic screening data, including in vitro validation. A robust software infrastructure was deployed to support upcoming project phases.
Finally, WP6 established the framework for an innovative clinical trial targeting five rare neuromuscular diseases. Two advisory structures—the Clinical Expert Committee (CEC) and the Patient Advisory Board (PAB)—actively contributed to defining clinical criteria and measurement tools. Three PRO tools (Patient-Reported Outcomes) are currently being evaluated through a patient survey. The consortium is also collaborating with other European initiatives to foster cross-project exchange on innovative trial methodologies and drug repurposing strategies.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

Currently, no technology is able to :
1. Quickly identify shared targets between diverse indications and
2. Efficiently identify drugs or combinations to modulate these targets.
On top of this, traditional drug discovery often relies on animal models or generic cell lines, which poorly mimic human pathophysiology.
DREAMS introduces an extremely novel paradigm that combines iPSC technology, high-throughput drug screening, and AI to identify shared pathological phenotypes, drugs that rescue them, and new therapeutic indications, to create a novel therapeutic discovery platform that goes beyond current methods. Our integrated approach enables AI-guided identification and testing of compounds directly on disease-relevant human cell types, increasing translational accuracy.

We have created artificial intelligence tools that combine information about gene activity with observable characteristics of organisms to accurately predict how effective and how safe different substances will be. The use of iPS-derived cells from the 5 rare neuromuscular disorders (NMD) that share common pathophysiological characteristics allows us to identify therapeutic responses in a personalized manner. This approach is expected to lead to the identification of several candidate molecules and novel targets, for rare diseases with currently limited treatment options.

DREAMS global project summmary