MELLODDY will demonstrate how the pharmaceutical industry can better leverage its data assets to virtualize the Drug Discovery (DD) process with world-leading Machine Learning (ML) technologies as an answer to the ever-increasing challenges and stricter regulatory requirements it is facing. The lack of a tested, secure and privacy-preserving platform for federated machine learning that enables pharmaceutical partners to extract DD-relevant information from all types of, not only their own but even each other’s competitive data, without mutual disclosure of the chemistry and biology each partner has worked on, has previously held back such demonstration, to the detriment of patients in the EU and beyond. MELLODDY’s ten pharmaceutical partners will enable this demonstration with an unprecedented volume of more than a billion highly private and competitive DD-relevant data points, and hundreds of Tbs of image data that annotate the biological effects of more than 10 million small molecules. The successful demonstration of the predictive benefits, i.e. increased predictive model performance and chemical applicability domain, of unlocking this data volume, while strictly preserving the privacy of all underlying data and the resulting predictive models, will shape best practices and translate into substantial efficiency gains in the DD process, and in the future, drug development. Finally, MELLODDY will prepare and exploit a service-for-fee vehicle to ensure the MELLODDY technologies are available to the rest of the pharmaceutical sector.

VICT3R, a public-private partnership running under the European Innovative Health Initiative, aims to significantly reduce the number of animals used in experimental studies performed during the nonclinical drug and chemical safety evaluation by replacing the animals of the concurrent controls groups (CCGs) with Virtual Control Groups (VCGs). These VCGs will be generated by means of state-of-the-art statistical or artificial intelligence (AI) approaches that optimally exploit the wealth of historical data from control animals accumulated over decades by pharmaceutical companies and other relevant industrial and academic sectors. The VCG concept was conceived and prototyped during the recently finished eTRANSAFE IMI2 project for its application in the nonclinical safety assessment of the pharmaceutical industry. A preliminary evaluation of the VCG concept carried out in the eTRANSAFE project demonstrated that it is generally feasible, yet scientifically and operationally challenging and must therefore be refined before its adoption for regulatory hazard and risk assessment. The main challenges consist of adequate data collection and curation, identification of key variables to achieve optimal matching between VCGs and CCGs, and validation of procedures including compliance with Good Laboratory Practice (GLP). These challenges will be systematically tackled in VICT3R for achieving the full development and regulatory acceptance of the VCG concept. VICT3R will collect, curate and analyse large data sets of control animals from different species to produce a large high-quality database. The database will be made available to VICT3R partners, regulators, and policy makers with the purpose to allow maturation of the VCG concept and to prove its validity, reproducibility and robustness. While VICT3R will be primarily focused on repeated dose toxicity studies, the extension of the VCG concept to other types of studies involving animals will also be tackled. VICT3R will promote that its database and software platform is maintained and expanded long term.