Figure 1

Pearson's correlation of real allele frequencies (x-axis) and frequency estimates from pooled DNA (RAS-scores, y-axis) of BACCS controls (a) (blood DNA) and the DeCC controls (b) (cheek swab DNA) are shown. The red arrows (a) indicate, that the distribution of RAS scores (y-axis) are now slumped towards 0.5 as a consequence of the discarded mismatch probe (thus mismatch probe intensities cannot be substracted from perfect match probe intensities) as compared to previous Affymetrix microarrays up to the 500 K. Pooled cheek swab DNA ( b ) could not sufficiently predict 'real' allele frequencies of the samples that comprised the pools.