Table 1.
Comparison of Average Error Rates of the PL Algorithm, the PGS Algorithm, and Clark’s Algorithm, for Two Real Data Sets
Average Error Rate (Standard Error) forb |
||
Algorithma | ACEc (L = 52) | CFTRd (L = 23) |
PL | .19 (.003) | .39 (.008) |
PGS | .36 (.004) | .48 (.009) |
Clark's | .27 (.000) | .47 (.018) |
The EM algorithm was excluded from the comparison because it cannot handle more than 15 heterozygous loci in the data.
Average error rates were defined as the number of erroneous phase calls divided by the total number of phase calls.
Average error rates were obtained by 100 independent runs of each algorithm.
Average error rates were for 100 data sets generated by randomly permuting 56 of the 57 complete haplotypes reported by Kerem et al. (1989).