Table 2.

Signatures of selection at MHC-I and MHC-IIB exon 2 of common buzzards

			Number of sites under selection		dN/dS and P for Z-test of positive selection (dN > dS)
MHC class	Exon	nseq (nAAsites)	Pos	Neg	All sites		PSS		Non-passerine PBR		Human PBR		Chicken PBR
					dN/dS	P	dN/dS	P	dN/dS	P	dN/dS	P	dN/dS	P
MHC-I	2	10(87)	4	6	0.713	1.000	3.336	0.004			3.293	0.007	2.579	0.013
MHC-I	3	91(87)	14	9	1.054	0.429	2.893	0.007	1.137	0.394	1.708	0.181	1.733	0.140
MHC-IIB	2	41(85)	16	3	1.337	0.171	11.472	1E-09	9.353	2E-04	7.457	0.001	5.152	2E-04
MHC-IIB (no recomb.)	2	36(85)	15	3	1.366	0.162	5.957	3E-10	6.622	2E-04	5.143	2E-05	5.058	5E-06

Strength of selection was measured with the number of sites under pervasive and episodic positive selection (as inferred with FUBAR and MEME algorithms, respectively) and with the ratio of nonsynonymous substitutions per nonsynonymous site to synonymous substitutions per synonymous site, averaging over all sequence pairs (dN/dS) and Z-test of positive selection (dN > dS) at: (i) all sites; (ii) positively selected sites (PSS); (iii) putative peptide-binding residues (PBR) of non-passerines, as inferred from the global analysis of selection at the avian MHC [34]; and from the crystallographic structure of MHC-I and II molecules for (iv) human peptide-binding residues [19, 20] and (v) chicken peptide-binding residues [23, 59]. Estimates were inferred for all MHC-I and IIB sequences, and for MHC-IIB exon 2 alleles with non-recombinant segments (no recomb.). n_seq: number of sequences, n_AAsites: number of amino acid sites. Significant (p < 0.05) ratios of dN/dS are in bold