Fig. 4

Landscapes of GC content, amino acid usage and protein evolution in mammals and reptiles. Genome-wide landscapes of base composition of predicted mRNAs correlate to the use of amino acids encoded by GC rich and GC depleted codons and to the rate of divergence of orthologous proteins. panels a-d data for two placental mammals: Odocoileus virginianus texanus (white-tailed deer, light blue line) and Panthera pardus (leopard, black line). panels e-h data for non-avian reptiles: Alligator mississippiensis (American alligator, dark blue line) and Chrysemys picta (painted turtle, red line). GARP% represents the sum of mole % of glycine, alanine, arginine and proline in the amino acid composition of the encoded proteins while FYMINK% represents the sum of mole % of phenylalanine, tyrosine, methionine, isoleucine, asparagine and lysine. The maxima and minima in the landscapes of protein divergence (100% minus protein identity) for the comparison between the two mammals (panel d) and the two reptiles (panel h) are strongly correlated to the landscapes of amino acid usage (R = 0.92 for GARP and R = -0.95 for FYMINK)