Mapping small molecule binding data to structural domains

BMC Bioinformatics

Table 5 Statistical analysis of power-law parameters

	Frequency of Pfam domains	Ligands per Pfam domain family	Ligands per target
xmin	10	81	210
alpha	2.07	1.71	2.15
Goodness of fit	0.5	0.42	0.42
vs_lognormal	yes (p = 5.1*10^-9)	yes/no (p = 0.48)	yes/no (p = 0.57)
vs_exponential	yes (p = 3.9*10^-3)	yes (p = 0.10)	yes (p = 8.5 *10^-8)
vs_weibull	yes (p = 2.1*10^-4)	Yes/no (p = 0.16)	no (p = 1.0*10^-3)
magnitude	~ 3	~ 3	~ 1
support for power-law	yes	yes	no

Parameters of the power-law functions fitted to the observed distributions of Pfam-A domain frequencies (left column), number of ligands associated with each Pfam-A domain (middle column) and number of ligands associated with individual targets (right column) are shown in columns 'xmin' and 'alpha'. 'Goodness of fit' indicates the p-Value calculated from a KS goodness of fit test. The rows vs_lognormal, vs_exponential, vs_weibull indicate outcomes of maximum-likelihood tests against alternative distributions. 'Yes' indicates significant support for a power-law distribution, 'no' indicates support for the alternative over a power-law. 'Magnitude' specifies the orders of magnitude in the distribution spanned by a power-law and 'support for power-law' is the summary outcome for each distribution.

ISSN: 1471-2105