The rates of sample exclusion due to missing genotypes are higher than usual due to technical issues during genotypic data collection. Degradation of a filter in a fluorescence reader reduced signal at the detector, thus artificially increasing the minimum signal necessary for detection. This attenuation of signal did not affect robustly performing samples, but did cause a higher fail rate among more weakly amplifying samples. Prior to exclusion of poorly genotyped samples, average genotyping completion by marker was 93.5% (90.5–96.4%); following exclusions, it was 97.3% (95.2–99.2%). Following identification of the problem and replacement of the filter, 4 SNPs (rs13393273, rs7597593, rs1344706 and rs4667001, including the two most strongly associated) were completely regenotyped. This repeated data collection yields a higher genotyping completion rate (mean 98.5%) and a low rate (0.52%) of discordant genotypes for these four markers.
