Figure 1

Reliable prediction of essential genes. Prediction of essential genes depends on both the quality of the network and the efficacy of the prediction algorithm to reproduce the features of essentiality. To account for these two factors, our method uses a set of n networks modelling a biological process, and m different centrality measures are applied to these to identify both the network(s) with the largest predictable set of essential genes and the centrality(ies) more effective to identify the essential genes. To determine the reliability of these predictions it is necessary to have access to the set of known essential genes.