Table 2.

Influence of expression breadth, phyletic age and expression abundance on protein properties using generalized linear model^a

Category	Property	P(phyletic age)	P(expression breadth)	P(expression abundance)
Structural
	Protein length (log)	< 10-15***	4.7 × 10-13***	< 10-15***
	Exon number	< 10-15***	< 10-15***	5.50 × 10-13***
	CpG+/TATA- promoter	2.30 × 10-14***	< 10-15***	0.0405*
	1st intron length (log)	0.0061**	0.0240 *	4.43 × 10-05***
	Length 5' UTR (log)	0.00747**	< 10-15***	4.31 × 10-05***
	GC of CDS	< 10-15***	< 10-15***	8.34 × 10-11***
Functional
	Molecular function (GO)	< 10-15***	< 10-15***	0.0739
	Pathway class (KEGG)	< 10-15***	7.17 × 10-6***	0.0121*
Evolutionary
	Ka	< 10-15***	< 10-15***	0.180
	Ks	7.84 × 10-5***	2.41 × 10-7***	0.428
	Ka/Ks	< 10-15***	< 10-15***	0.000354***
MainFactors
	Age	-	< 10-15***	0.348
	EST breadth	< 10-15***	-	< 10-15***
	Expression abundance	0.348	< 10-15***	-

Numbers are the corresponding P-values, * P < 0.05;** P < 0.01;*** P < 0.001.

^a We used a form like 'property ~ phyletic age + expression breadth + expression abundance' in the generalized linear model analysis. This form will produce three p-values showing the influence of phyletic age, expression breadth and expression abundance on 'property' respectively; p-values less than certain threshold (0.05 for example) suggest significant contribution of some factors to the 'property'; multiple significant p-values suggest the corresponding factors contribute independently to the 'property'.