. 2009 Feb 23;4(2):e4549. doi: 10.1371/journal.pone.0004549

Table 4. Frequency spectra in coding (C) and non-coding (NC) regions across species at each gene.

Population	Region	Def [C:306/NC:978–1016 nt]				SP14D [C:1083/NC:588–607 nt]				Gambic [C:243/NC:415–432 nt]				GNBP [C:1188/NC:930–959 nt]
		f = 0^a	f = 1^a	f = 2–3	f = 4–7	f = 0	f = 1	f = 2–3	f = 4–7	f = 0	f = 1	f = 2–3	f = 4–7	f = 0	f = 1	f = 2–3	f = 4–7
A. gambiae	Coding	96.7	0**	1.3*	2	97	2	0.7	0.7***	97.9	1.7*	0.4	na	96.1	3	0.6*	0.3
West Kenya	NoCod	91.4	3.1	4.2	1.3	90	3.4	1.7	5.2***	91.9	6.7	1.4	na	92.2	4.7	2.5**	0.6
A. arabiens	Coding	95.8	2	1.3	1	97	1.7*	0.7**	0.3	97.1	1.7	0.8	0.4	97.2	2.1	0.5	0.2
	NoCod	94.3	3.1	1.3	1.3	90	4.8**	3.8***	1.2	94	0.7	3.4	1.9	95.9	2.4	1.7*	0
A. quad	Coding	98	1.3*	0.3	0.3	97	1.9	0.4	0.6	98.4	1.2	0*	0.4	97.1	2.1	0.7	0.2
	NoCod	91.6	4.6	1.9	1.9	94	4.1*	1.3	1.1	93.4	2.3	3.5	0.5	95.4	3.4	0.7	0.4
All (Pooled)	Coding	96.1	1.4***	1.3**	1.3	97	2***	0.6***	0.4***	97.9	1.3**	0.5***	0.3	96.8	2.5	0.5***	0.2*
	NoCod	92	3.9*	2.6	1.5	91.4*	4.1***	2.4***	2.1***	93.5	3.6*	2.3*	0.7	94.5	3.1	1.8***	0.6*
	Overall	93	3.3	2.3	1.5	96	2.7	1.2	0.5	95.1	2.8	1.6	0.5	95.6	2.7	1.1	0.4

Frequency spectra classes including invariant positions (f = 0), low polymorphism represented by singletons (f = 1), moderately polymorphic positions with the rare nucleotide observed twice or three times (f = 2–3), and highly polymorphic positions with the rare nucleotide observed four or more times (f = 4–7). The relative distribution of each class is expressed as percentages. Excess and deficit of observed vs. expected frequency is marked by red and blue respectively in cells with significant deviations based on 1 df χ² test (^*, ^**, ^***, represent P<0.05, 0.01, and 0.001, respectively). The western Kenya population of A. gambiae represents this species (heterogeneity χ² test showed no evidence for heterogeneity among the four populations). All contingency tables for each gene and species were significant (P<0.01).