Figure 1

Information content versus the number of tag SNPs in the ENr112 region from CEU population. (a) The information content is computed according to the model used to select the tag SNPs. Top-down: Markov model (dotted black line), unconstrained and Daly et al. HMMs (dot-dashed gray and black lines, respectively), models with context sizes 1 and 2 (dashed gray and black lines, resp.), homogeneous and heterogeneous Li and Stephens HMMs (solid gray and black lines, resp.). (b) The information content is always computed according the heterogeneous Li and Stephens model, regardless of the tag SNP selection method. The symbols for the tag SNP selection methods are the same as in (a). Additional symbols show the information content of evenly spaced tag SNPs (black circles), and randomly chosen tag SNPs (gray circles).