The MechML project is a BMBF initiative that aims to  bridge the gap between machine learning methodology and challenging life sciences applications by exploiting biological knowledge to allow effective, interpretable and scalable learning.

The project aims in extending existing approaches of ML and computational biology to meet relevant requirements with new methods. This includes

(i) improving the interpretability of highly parameterized multi-layer models (deep learning),

(ii) combining conventional approaches of ML with mechanistic models and expert knowledge,

(iii) developing models that enable robust learning with existing statistical dependencies between observations and disruptive factors,

(iv) developing transferable models and software systems to apply models to new data and different technologies, and

(v) developing software systems for scalable and parallel inferences.