According to the experimental results of LAC algorithm, this algorithm is superior to the six typical centrality measures
mentioned above. In addition, LAC considers the important modular nature of protein essentiality. Hence, we use the dynamic local average connectivity to represent the topology centrality of proteins. Meanwhile, Hart et al. (2007) pointed out that protein complexes highly correlate with essential proteins. Therefore, we choose the dynamic complex centrality as the biological part of our essentiality measure. Based on the analysis above, we select the integration of the dynamic local average connectivity and complex centrality to form our final essentiality measure of proteins. To combine these two parts, a harmonic centrality is defined as Function (6).