LAUSR

When combined with results from the rs73885319 and rs71785313 SNP assays, will result in an inferred haplotype not observed on African chromosomes. To the Editor: There are 2 APOL1 variant haplotypes, G1 and G2, that predispose individuals of African descent to a diverse spectrum of chronic kidney disease in the homozygous (G1/G1 or G2/G2) or compound heterozygous state (G1/G2). The G1 haplotype (approximately 23% of African Americans) carries 2 nonsynonymous variants: rs73885319 (p.S342G) and rs60910145 (p.I384M). Genotyping rs73885319 is generally sufficient, as the 2 singlenucleotide polymorphisms (SNPs) are in near-absolute linkage disequilibrium; however, researchers often genotype both G1 variants for quality control purposes and to distinguish between 2 G1 sub-haplotypes: G1 (has both S242G and I384M amino acid substitutions) and the rare G1Gþ (only the S242G substitution). The G2 haplotype harbors a 6–base pair in-frame deletion (rs71785313, also known as rs143830837, approximate 13% frequency in African Americans). We use custom Taqman assays for these SNPs: rs73885319 (Assay ID-AH20SD1), rs60910145 (Assay ID-AHWR1JA), and RS71785313 (Assay ID-AH1RT7T). Many laboratories type the APOL1 SNPs using Taqman assays that are available from ThermoFisher Scientific (Waltham, MA) as predesigned assays. The ThermoFisher Web site notes that one of the polymerase chain reaction primers for rs60910145 (Assay ID-C__89555688_10) spans the 6–base pair deletion of the G2 SNP. We found that the presence of the 6–base pair deletion results in a failure of amplification of the rs60910145 allele on the G2 haplotype. Thus, the G2/G2 haplotype will fail to amplify for rs60910145, and G1/G2 individuals will fail to amplify the rs60910145 allele on the G2 haplotype and be scored as homozygous for the variant nucleotide instead of heterozygous (Table 1). Although the G2 SNP is absent or rare in Asian and white individuals, and thus will not affect RS60910145 typing in those populations, most investigators are typing RS60910145 in populations of African descent in which G2 is much more common. We have typed a set of 1110 African American samples with the 2 different versions of the rs60910145 assay, as well as rs73885319 and rs143830837, and confirmed this result (7.3% of the rs60910145 genotypes obtained with the predesigned assay failed or gave the incorrect genotype). Specifically,