Supplementary material for Hassanin et al. (2000) Proc. Natl. Acad. Sci. USA 97 (21), 11415-11420.

To measure independently the levels of homoplasy for each class of substitutions (i.e., C-T, A-G, A-C, A-T, C-G, and G-T) at each of the three codon positions, the DNA alignment of exon 2 used for phylogenetic analyses was recoded into 18 binary data subsets (1). Each binary data subset (subsets correspond to one type of substitution at one codon position) was then analyzed by using PAUP 3.1.1 (2), and two different approaches were used to quantify its homoplasy content (e.g. C-T substitutions at third codon positions; see supplemental Fig. 6) (1). The first is based on the CI excluding uninformative characters, which measures the amount of homoplasy. CI corresponds to the minimum number of steps required by the data divided by the number of steps in the most parsimonious tree. It is standardized to be no more than 1 (no homoplasy) and no less than 0. The second approach is based on the analysis of saturation, which describes graphically the distribution of homoplasy within the sequence sample. One of the most-parsimonious trees was used as a basis for assessing graphically the saturation of the substitution type under focus. By using the patristic distance matrix for binary characters (patristic distances option of PAUP), the pairwise number of observed differences (i.e., "adjusted character distances" of PAUP) was plotted against the corresponding pairwise number of inferred substitutions occurring along branches joining the two V_H sequences in the most-parsimonious tree (i.e., "patristic distance" of PAUP). For each substitution type, the S was used to estimate the level of saturation. When no saturation is observed, the slope of the linear regression equals one. When the level of saturation increases, the slope decreases towards zero.

1. Hassanin, A., Lecointre, G. & Tillier, S. (1998) C. R. Acad. Sci. Ser. III 321, 611-620.

2. Swofford, D. L. (1993) PAUP: Phylogenetic Analysis Using Parsimony (Illinois Natural History Survey, Champaign, IL), Ver. 3.1.1.