< Back to Article

Linkage Disequilibrium-Based Quality Control for Large-Scale Genetic Studies

Figure 4

Calibration of conditional probabilities of corrected genotypes in the Affymetrix data.

Separately in 3 HapMap populations and for all 22 autosomes, we calculated the conditional probabilities of genotypes other than the observed genotype calls. We then binned these probabilities and, within each bin, calculated the proportion of genotypes which, if switched to the most-probable genotype other than that observed, resulted in a decrease in the number of discrepancies with the HapMap calls.

Figure 4