All patients underwent genotyping when they enrolled in the Twin-Sibling Study, which was performed in two phases as a part of the International CF Modifier Consortium23 (link). Genotyping was carried out on the Illumina Quad 610 platform for Phase 1 and on the Illumina 660W and Omni5 platforms for Phase 2. MaCH/Minimac software was used for calculating genotypes based on the genotyped data. For calculation, the reference used was Phase I, Version 3 haplotype data from the 1000 Genomes project24 (link), including all reference samples from the 1000 Genomes project.
At the TAS2R38 locus, we identified rs713598, rs1726866, and rs10246939 in codons 49, 262, and 296, respectively. rs713598 was imputed and rs7126866 and rs10246939 were genotyped on all platforms. The imputation quality for rs713598 in Illumina Quad 610, Illumina 660W, and Omni5 platforms were 0.93, 0.90, and 0.93, respectively, demonstrating that the imputation was of high quality. Next, the genotyping was scored based on the negative strand and coded, such that patients with (CC) (CC) (GG) nucleotide pairs at the respective SNP locations were designated homozygous for the functional haplotype (PAV/PAV). Patients with (C-) (C-) (G-) were designated heterozygous for the functional haplotype (PAV/*), and those without this coding were designated homozygous for the non-functional allele.