On the automated function prediction of big multi-species networks

Automated Function Prediction (AFP) of proteins in Multi-species Networks raises challenging computational problems due to the size of the generated network and the lack of scalability of traditional approaches. We present a network-based approach that allows to predict protein functions in a multi-species setting by exploiting homology relationships between species. The method adopts secondary memory-based technologies to efficiently process huge protein networks using ordinary stand-alone machines.