ELIXIR is tackling a novel discipline by establishing a new ELIXIR Food and Nutrition Community. This Community  brings together experts from across Europe who will drive the development of new guidelines and standards for researchers and developers who are providing solutions to a broad spectrum of problems. 

Led by Jildau Bouwmann from the Dutch Techcentre for Life Sciences (ELIXIR-NL), the Community will strive to improve preventive health concerning food and nutrition, and use the consumer as a linking pin for supporting lifestyle interventions and healthy food patterns. Visit the ELIXIR Food and Nutrition Community page to read more about the expected outcomes, and how to join the Community.

ELIXIR has launched the Biocuration Focus Group in collaboration with the International Society for Biocuration (ISB). This Focus Group will widen the applicability of ELIXIR activities to data resource developers/operators across the Nodes, and  act as a connection point for all of the ELIXIR members and their close collaborators who have biocuration as their main area of responsibility.

Guidelines published in Nature Methods, to ensure quality and reproducibility of predictive methods

An international group of scientists, including the ELIXIR Machine Learning Focus Group, have published a set of guidelines for better reporting standards for AI methods that seek to classify biomedical data. The recommendations are published in Nature Methods. Tom Lenaerts, member of the ELIXIR Machine Learning Focus Group and current director of IB², expressed the need for such guidelines, stating:
 

"The popularity of machine learning and deep learning today gives the impression that new AI tools can be designed quickly without thinking much about the data and the actual objectives. That is not the case. Inaccuracy is easily achieved when one does not fully understand the nature of the data and the characteristics used in a forecasting method. In the medical and biological fields, more than half the time should actually be spent designing a high-quality data set and finding the right set of characteristics to train the method ."