Figure 3.
Sequence similarity between PLS or LSP2 in the sequence database containing all the PCMPs. HMMs have been built up with 20 amino acid block sequences from either position-1 PLS (A), position-1 LSP (= LSP2; B), or position-2 PLS (C). The PLS block at position 1 is the first PLS block on the N-terminal side of the P2L2S2 (or PL2S) block in the proteins (Table I; Fig. 1). The PLS block at position 2 is the next one toward the N terminus of the protein and so on for position 3 and others. Hmmsearch output, sorted by increasing E-value, has been organized in classes of 20 PCMP blocks as shown for A in the insert at the top right corner. E-value classes, illustrated by a bar, are ordered along the abscissa by increasing E-values. The E-value class of rank 1 (E-value class 1), contains the 20 PCMP blocks showing the highest similarity with the HMM and the similarity decreases with increased E-value class ranks. For other sequence comparisons, in B and C, only the highest and lowest E-values are given. Different patterns in bars indicate the numbers of PCMP blocks that are located at different protein positions: black for blocks at position 1, white for blocks at position 2, dark gray for blocks at position 3, and light gray for those at other positions. The number of blocks at the bottom of each bar (or E-value classes) is always for the blocks belonging to the same category as the 20 blocks used to build up the HMM. A regression line has been calculated for the number of these blocks. The line has been forced to horizontal when the slope was not significant. The stronger the slope the higher is the similarity of blocks belonging to the same category as the 20 blocks used for building the HMM (A) or the distance with blocks of other categories (B).