Completion of biological networks: the output kernel trees approach

The inference of biological networks from various sources of experimental data is an important problem of computational biology. In this paper, we propose a new method for the supervised inference of biological networks, which is based on a kernelization of the output of regression trees. It inherits several features of this method such as interpretability, robustness to irrelevant variables, and input scalability. We applied this method on the inference of a protein-protein interaction network where we obtained results competitive with existing approaches. Furthermore, our method provides relevant insights on input data regarding their potential relationship with the existence of interactions.