Supplementary data 2

Phylogenetic tree of the 47 MhcSF proteins inferred using PHYML maximum likelihood program [1] with WAG substitution model, rooted with a MHC-II protein sequence (Homo sapiens HLA-DM) and drawn with NJplot program [2]. Each protein belongs to one receptor type among 9 (MHC-I, AZGP1, CD1, EPCR, FCGRT, HFE, MIC, MR1 and RAE) and to one species among 4 (Hs: Homo sapiens, Mm: Mus musculus, Rn: Rattus norvegicus, Bt: Bos taurus). The 47 alleles (one for each protein) used are described in Supplementary data 1, and their aligned sequences are available in Protein sequence alignment (IMGT web site). Full and dotted vertical lines indicate B2M bound and unboud receptor types, respectively. 1000 replicates were used to estimate bootstrap supports.

[1] Guindon,S. and Gascuel,O. (2003) A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst. Biol., 52, 696-704.

[2] Perrière, G. and Gouy,M. (1996) WWW-Query: An on-line retrieval system for biological sequence banks. Biochimie, 78, 364-369.