Our organisation

Entelos is a non-profit research institute providing advanced data management, computational tools, and predictive models to promote data-driven innovation and address various global challenges. Our research includes, among others, data management practices for R&D, drug discovery, materials design, hazard, and risk prediction. The Entelos mission is to make these tools accessible to everyone, regardless of their background or location. To do so, Entelos is developing easy-to-use graphical interfaces enabling users to run models and understand the outputs. Entelos is committed to improve scientific data stewardship and promote open access to high-quality curated data and associated metadata, supporting collaboration across different fields of research, and enhancing the impact of the data for policy, regulations, and decision-making. Our aim is to advance the EU digital single market and provide positive socioeconomic impact for the EU from publicly funded data.  In its pivotal role within the SSbD4CheM project, Entelos is set to harness its specialized capabilities in the domains of Metadata Management and Data-Driven Modelling, bringing a deeply technical and analytical approach. In Metadata Management, Entelos’ expertise lies in the intricate orchestration of data ontology, taxonomy, and schema development. This involves constructing a comprehensive framework for data classification, indexing, and storage, which is critical for ensuring the integrity and accessibility of complex, multidimensional datasets. The focus is on establishing robust data governance protocols and ensuring semantic consistency across diverse data sources, thereby facilitating interoperability and efficient data retrieval. On the front of Data-Driven Modelling, Entelos employs sophisticated algorithmic strategies and machine learning techniques to mine and interpret this structured data. The approach encompasses the development of statistical models and computational algorithms tailored to discern underlying patterns and relationships within the data. This includes the application of regression analyses, cluster analyses, and predictive modelling to extract meaningful insights from chemical/ (nano)materials and biological datasets. The objective is to transform raw data into a coherent narrative that can predict materials behaviours, simulate biological interactions, and guide safe by design decisions. 

Our role in the project

Entelos Research Institute will apply Data-driven modelling to predict the properties of candidate materials quickly and reliably. Entelos will use the data produced from muticale simulations to develop initial data-driven workflows to predict the properties and behaviour of materials in different environments based on state-of-the-art algorithms (e.g., J48, Random Forest, k-nearest neighbours). These will be complemented, if possible, with interoperable and reusable literature data following rigorous quality control (e.g., completeness, scientific validity). The datasets and associated workflows produced by Entelos will be used to design the minimum number of experiments required to fill the critical data gaps that will be fed to the experimental part of the project.

Main person involved

Lefteris Zacharia
Data driven modelling