Figure 1.
Allele frequency spectra for different classes of markers.
Note that the scale on the x axis is different for total SNPs, because only this class is biallelic.
Table 1.
Properties of different marker sets.
Table 2.
Number of haplotypes observed in multi-SNP amplicons.
Figure 2.
Estimated ln(probability of the data) and Var[ln(probability of the data)] for k from 2 to 5.
Values are from STRUCTURE run three times at each value of k, using A) 89 SSRs; B) 847 SNPs; C) 554 SNP haplotypes; D) 89 SSRs+847 SNPs. The blue diamonds are ln(probability of the data) and the pink squares are var[ln(probability of the data)].
Table 3.
Percent population assignment based on different marker sets.
Figure 3.
Comparison of membership in population clusters based on marker class.
Each point represents one individual's proportion of ancestry in the NSS (top panel) or TS (bottom panel) cluster, based on SSR (x axis) or SNP (y axis) data assuming three populations. The arrows indicate line F2834T (see text).
Figure 4.
Correlation between genetic distance based on SSRs and SNPs.
Each point represents the genetic distance between a pair of individuals, based on sharing of SSR (x axis) or SNP (y axis) alleles. The top panel shows the relationship for pairs of individuals whose SSR distance is <0.65; the bottom panel shows the relationship for pairs of individuals whose SSR distance is ≥0.65.
Table 4.
Correlation between genetic distance for SSRs and SNPs of different frequency classes.
Table 5.
Alleles captured by subsets of the data for different marker sets.
Table 6.
Properties of data sets simulating full resequencing.