Characterizing Gene Expressions Based on Their Temporal Observations

Table 2

Covariance structures using ML.

Model	Description	AIC	BIC	−2log likelihood
1	General linear model (GLM)	1811.8	1856.2	1798.8
2	Compound symmetry (CS)	1811.5	1856.0	1796.7
3	Variance components (VC)	1665.0	1651.3	1645.0
4	Heterogeneous CS (CSH)	1636.8	1618.0	1610.8
5	Spatial power (SP)	1689.2	1685.6	1600.2

AIC: Akaike’s information criteria; BIC: Bayesian information criteria for each model selected.