A novel representation of RNA secondary structure based on element-contact graphs

BMC Bioinformatics

Table 2 Correlations between topological indices and their fitting models. Four typical distribution models (normal distribution, Gamma distribution, Weibull distribution and log-normal distribution) are employed here to model the statistical distributions of the three topological index families. The parameters of these distribution models are estimated through maximum likelihood method, and the goodness of model fitting is evaluated by Pearson's correlation coefficient. For each topological index, the highest Pearson's correlation coefficient is in bold.

Index family	Indices	Normal	Gamma	Weilbull	Log-normal
Wiener	$W_{S L}^{w}$	0.96	0.95	0.97	0.88
	W _SL	0.83	0.93	0.93	0.83
	$W_{S}^{w}$	0.83	1.00	0.99	0.96
	W _S	0.77	0.93	0.93	0.92
	$W_{L}^{w}$	0.93	0.94	0.95	0.85
	W _L	0.74	0.85	0.85	0.81
Balaban	$J_{S L}^{w}$	0.91	0.99	0.99	0.95
	J _SL	0.76	0.91	0.91	0.89
	$J_{S}^{w}$	0.82	0.98	0.96	0.93
	J _S	0.81	0.91	0.89	0.89
	$J_{L}^{w}$	0.92	0.93	0.94	0.86
	J _L	0.64	0.84	0.84	0.82
Randić	$\begin{matrix} 0 \end{matrix} χ_{S L}^{w}$	0.93	0.80	0.86	0.69
	$\begin{matrix} 1 \end{matrix} χ_{S L}^{w}$	0.84	0.93	0.93	0.88
	⁰ χ _SL	0.72	0.74	0.75	0.69
	¹ χ _SL	0.82	0.79	0.81	0.71
	$\begin{matrix} 0 \end{matrix} χ_{S}^{w}$	0.95	0.96	0.96	0.89
	$\begin{matrix} 1 \end{matrix} χ_{S}^{w}$	0.82	0.99	0.98	0.94
	⁰ χ _S	0.57	0.57	0.58	0.51
	¹ χ _S	0.87	0.91	0.91	0.86
	$\begin{matrix} 0 \end{matrix} χ_{L}^{w}$	0.92	0.96	0.97	0.87
	$\begin{matrix} 1 \end{matrix} χ_{L}^{w}$	0.80	0.99	0.99	0.95
	⁰ χ _L	0.52	0.53	0.54	0.51
	¹ χ _L	0.81	0.85	0.86	0.80

ISSN: 1471-2105