Marc Joired and Liesbet Geris have published the paper ‘Cracking the genetic code with neural networks’ in Frontiers in Artificial Intelligence. It is a didactical/pedagogical paper targeting the biomedical community about the trade-off between ‘understandability/complexity’ of a deep learning algorithm and data efficiency (amount of data required to reach a given loss or a given accuracy). The data are human transcriptomic/proteomic data pairs.

Read the paper here.