Table 1.

Graph properties calculated for a typical tRNA depicted in Figure 1

Number of articulation points	3	Average Burt's  constraint	0.4234
Average path length	9.577	Variance of Burt's constraint	0.0161
Average vertex betweenness	313.1	Average degree	2.514
Variance of vertex betweenness	54817.6	Diameter	22
Average edge betweenness	278.2	Girth	4
Variance of edge betweenness	40784.5	Average coreness	1.959
Average cocitation coupling	0.0555	Variance of coreness	0.0394
Average bibliographic coupling	0.0555	Maximum coreness	2
Average closeness centrality  index	0.1088	Graph density	0.0344
Variance of closeness  centrality index	0.00048	Transitivity	0

We have chosen 20 graph properties to calculate and train the SVMs. These properties were chosen by considering the following criteria (i) polynomial-time computation, (ii) relevance to local and global levels of the graph and (iii) usage in complex network research. The values shown beside each property are the graph properties as calculated for the graph shown in Figure 1.