TABLE 8.
Subject (i) | Quantitative phenotype |
Diplotype configuration |
True or falsea | Posterior distributionb |
Posterior distributionc |
---|---|---|---|---|---|
1 | 157.1 | GCCT GCCT | True | 1.0000 | 1.0000 |
2 | 170.3 | CCTC CTCC | True | 0.9993 | 0.9999 |
CCCC CTTC | False | 0.0007 | 0.0001 | ||
3 | 173.4 | CCTC GCCT | True | 1.0000 | 1.0000 |
4 | 158.2 | CTCT GCCT | True | 1.0000 | 1.0000 |
5 | 170.6 | CTCT GCCT | True | 1.0000 | 1.0000 |
6 | 162.4 | CTCC CTCC | True | 1.0000 | 1.0000 |
7 | 161.6 | CTCC GCCT | True | 0.9975 | 0.9975 |
CTCT GCCC | False | 0.0025 | 0.0025 | ||
8 | 149.4 | CTCC CTCC | True | 1.0000 | 1.0000 |
9 | 164.3 | CCTC GCCT | True | 1.0000 | 1.0000 |
10 | 165.1 | CTCC GCCT | True | 0.9975 | 0.9975 |
CTCT GCCC | False | 0.0025 | 0.0025 |
Simulations were started by assigning diplotype configurations to N = 1000 number of subjects according to the haplotype frequencies employed from the four-locus data for the SAA1 gene. Depending on whether the subject possessed the phenotype-associated haplotype CCTC, a quantitative phenotype was drawn from N(μ1, σ2) or N(μ2, σ2), where μ1 = 165, μ2 = 160, and σ = 5. After removing the phase information, QTLHAPLO was used to determine the posterior probability distribution of the diplotype configuration (diplotype distribution) for each subject either by using only genotype data or by using both genotype and phenotype data.
Possible diplotype configurations for each subject were compared with the diplotype configuration before the phase information was removed. “True” means that the diplotype configuration before the removal of the phase information was the same as the estimated diplotype configuration, while “False” means that they were different.
Posterior probability distribution of the diplotype configuration given only the observed genotype data [P(di = ak|Gobs)].
Posterior probability distribution of the diplotype configuration given the observed genotype and phenotype data [P(di = ak|Gobs, Ψobs)].