Estimating the cluster significance threshold given a user-defined p-value. An illustrating example is shown in which nr = 5 random data are generated, the data are subsequently clustered according the proposed clustering/selection procedure and cluster significance distribution are depicted in (a) and (b) following sorting. The corresponding p-value for each cluster significance cs is estimated and depicted in (c). Thus, given a p-value, we can infer the corresponding cluster significance threshold. For example, for a p-value = 0.05, all clusters with cluster significance ≥ 10 are selected and if p-value = 0.1, all clusters with cluster significance ≥ 8 are considered as significant clusters.