The main objective will be to construct appropriate mathematical models for each type of food enabling the proper evaluation and prediction of the impact of variables linked with future changes on the fermentation processes. It is highly interwoven with three experimental subtasks planed in Task 4. Screening factorial models will be designed to test a large number of variables (subtask 3.1) and predictive models to deeply characterize and predict the impact of selected variables (subtask 3.2). Finally, multi-omics data integration will be used to construct ecological networks (subtask 3.3).

> 3.1. SCREENING FACTORIAL MODELS

> 3.2. PREDICTIVE MODELS

> 3.3. ECOLOGICAL NETWORKS