Table 2. Variations in (C+G)% and ρ(CG) statistics among differing RNA classes.

RNA	Species	(C+G)%	ρ(CG)
tRNAs (cytoplasmic)	M.jannaschii	66.2	0.72
	Plasmodium	25.6	0.88
	C.elegans	58.8	0.92
	H.sapiens	58.0	1.00
rRNAs	M.jannaschii	63.8	0.87
	Plasmodium	36.0	0.98
	C.elegans	48.0	1.04
	H.sapiens	60.3	1.03
SRP RNAs	M.jannaschii	65.7	0.66
	C.elegans	56.8	1.07
	H.sapiens	58.3	0.48
Small nuclear RNAs	C.elegans	43.1	1.01
	H.sapiens	44.3	0.90
sno- and sno-like RNAs	M.jannaschii	48.7	0.83
	H.sapiens	44.2	0.38
Riboregulator RNAs	H.sapiens	47.6	0.33
	All eukaryotes	44.4	0.46
Pseudoknot RNAs	All bacteria	52.0	0.86
	All viruses	51.8	0.98

This table shows some of the variations in (C+G)% and ρ(CG) values among different classes of RNAs for M.jannaschii, Plasmodium (multiple species combined because of limited data), C.elegans and H.sapiens. For the riboregulator and pseudoknot databases, which contain limited or no data for these species, RNA sequence data from other groups of organisms are shown. The base-composition variations shown in the table indicate which ncRNA types are more likely to be detected by a RNA gene-finder based on (C+G)% and ρ(CG). For example, one notes that (C+G)% is only 48.7% in M.jannaschii snoRNAs and 43.1% in C.elegans snRNAs. Consequently these RNAs would be more difficult to detect using a (C+G)% detector than the more G+C rich tRNAs, rRNAs and SRP RNAs.