Conservation of the gene batteries. (a) Conservation profiles of the gene batteries. For each battery, the over-represented motif(s) found in the reference sequence is assessed for over-representation in the corresponding regions of the homologous target genes in 24 other eukaryotic species. The reference species is the one from which the ChIP data were collected (H. sapiens or M. musculus if labeled with an asterisk). In red are the species whose over-representation score is above 8; in black are the species whose over-representation scores are between 4 and 8; and in blue are the species whose over-representation scores are lower than 4. The higher the over-representation score, the more over-represented is the motif in that species and, hence, the more conserved is the network compared to the reference species network. A significant over-representation score is 4 or above (see Material and methods). The values in parentheses correspond to the number of genes forming the batteries in the reference species. (b) Conservation profiles of the regulatory networks using non-alignable sequences: same as (a) except that the sequences used have been masked in the region where a significant alignment can be found with the reference sequence. Grey boxes correspond to the reference species, which, by definition, does not have unaligned sequences. For numerical values, see Additional data file 4. (c) Pie chart representing the variable degree of conservation of the various gene batteries analyzed: 1 (6%) gene battery is conserved through the primate lineage (NfKb); 5 (31%) are conserved in most mammals (SRF, POU5F1, SOX2, HNF1A, HNF4A); 4 (25%) are conserved through the tetrapode lineage (Myod1, Myog, NRF1, HNF6 (ONECUT1)); 5 (31%) are conserved through the vertebrate lineage (YY1, ETS, CREB1, ESR1, NOTCH1); and only 1 (E2F) is conserved through the metaozan lineage.

Assessment of the ancestral cores. For each gene battery we show the probability of the genes to be part of the ancestral core for lineage b given that the genes are part (blue) or not (green) of the ancestral core of lineage a. Significant differences between P(core b | core a) and P(core b | not in core a) are indicated by asterisks (p-values < 0.001). Three phylogenetic distances were considered: (a) mammalian; (b) vertebrates; (c) chordates.