Table 1. Structural properties and ranking results of different methods in real networks.

Network	Network properties				Random scheme				Local scheme
	N	〈k〉	D	λc	〈τ〉d	〈τ〉b	〈τ〉k	〈τ〉RLP	〈τ〉LD	〈τ〉RLP
Dolphins	62	5.13	8	0.172	0.776	0.531	0.775	0.830	0.642	0.757
Word	112	7.59	5	0.078	0.764	0.639	0.754	0.815	0.758	0.821
Jazz	198	27.70	6	0.027	0.665	0.519	0.671	0.791	0.633	0.835
E. coli	230	6.04	11	0.075	0.713	0.491	0.752	0.840	0.690	0.833
C. elegans	297	14.46	5	0.040	0.687	0.577	0.700	0.780	0.665	0.780
Netsci	379	4.82	17	0.142	0.443	0.305	0.415	0.803	0.629	0.799
Email	1133	9.62	8	0.057	0.759	0.637	0.775	0.799	0.718	0.777
Blog	1222	27.36	8	0.013	0.708	0.607	0.713	0.724	0.782	0.792
TAP	1373	9.95	12	0.065	0.675	0.352	0.669	0.824	0.526	0.733
Y2H	1458	2.67	19	0.163	0.301	0.289	0.346	0.632	0.586	0.791
HEP	5835	4.73	19	0.123	0.534	0.403	0.562	0.641	0.579	0.708
PGP	10680	4.55	24	0.056	0.494	0.357	0.509	0.728	0.517	0.736

Structural properties include network size (N), average degree (〈k〉), network diameter (D), critical infection probability λ_c . The random scheme represents the case where 10% nodes are set as the target nodes and randomly distributed in the network. The local scheme stands for the case where 10% nodes are set as the target nodes and locate in the nodes with maximum distance L measured by the shortest path length to a randomly selected central node (L = 3 in PGP, L = 4 in HEP, L = 8 in Y2H and L = 2 in the rest of networks). According to Fig. 4, we compare τ of four methods including degree (τ_d), betweenness (τ_b), k-core (τ_k) and RLP (τ_RLP) in the random scheme. According to Fig. 5, we compare τ of two methods including Local degree (τ_LD) and RLP (τ_RLP). The infection probability for the SIR model in each network is set as λ_c. The results of the RLP method are highlighted.