Figure 1
From: A novel series of compositionally biased substitution matrices for comparing Plasmodium proteins
Amino acid deviations from BLOSUM62 target frequencies. We generated two datasets from each of the proteomes of P. falciparum and P. yoelii. One dataset contained proteins for which an annotated ortholog existed in a non-Plasmodium species (PfO, PyO), and the other contained proteins classified as hypothetical and which were not in the first dataset (PfH, PyH). The amino acid background frequencies implicit in the BLOSUM62 matrix were derived. P. falciparum and P. yoelii proteins in the four data sets PfO, PyO, PfH and PyH were filtered using the SEG algorithm (window size 24; locut 2.3; hicut 2.4), then for each data set the percent divergence of each amino acid from the BLOSUM62 background was calculated.