Genomic data assimilation using a higher moment filtering technique for restoration of gene regulatory networks

Table 4 The comparison results using the original model given by GeneNet and the five time-courses generated from a yeast cell-cycle network with Gaussian system noise of mean 0 and variance 0.01

	PR	RR	F-measure	TP	FP	TN	FN
HMEnPF	0.827	0.738	0.778	19.2	4.2	59.8	6.8
UKF	0.703	0.708	0.704	18.4	8.0	56.0	7.6
GeneNet	0.413	0.520	0.460	10.4	14.8	10.2	9.6

The comparison results of the proposed algorithm, the previous algorithm using UKF only [36], and GeneNet for the five time-courses generated from a yeast cell-cycle network with Gaussian system noise of mean 0 and variance 0.01. The results of ‘PR’, ‘RR’, ‘F-measure’, ‘TP’, ‘FP’, ‘TN’ and ‘FN’ for the five time-courses were averaged. The networks inferred by GeneNet were used as the original networks for the former two algorithms.