| Long-range PCR assays were designed using Oligo primer design software |
| (Molecular Biology Insights). Primers were selected to have similar |
| stringency and to map uniquely to NCBI Build 34. From a collection of |
| all suitable candidate primers, we used custom software to select a |
| minimum spanning set having maximum coverage with minimal overlap |
| between adjacent amplicons. |
| Oligonucleotide arrays were designed so each SNP would be interrogated |
| by forty distinct 25 bp probes. These consisted of four sets of 10 |
| features, corresponding to forward and reverse tilings of sequences |
| complementary to each SNP allele. A set of 10 features consisted of |
| five sets of two features where the location of the SNP within the |
| oligonucleotide varies from position 11 to position 15. For each |
| offset, we tiled one perfect-match feature and one feature having a |
| mismatch at the central position of the probe. Thus, for each allele, |
| we tiled ten perfect-match and ten mismatch probes. Mismatch probes |
| were used to measure background, and by comparison with the signal for |
| the perfect match probes, to detect the presence or absence of a |
| specific PCR product. |
| Light-directed chemical synthesis of oligonucleotide arrays was |
| carried out by Affymetrix, Inc. (Santa Clara, CA). Genomic DNA was |
| amplified by multiplex long-range PCR, then all PCR products destined |
| for a single genotyping array were pooled together and end-labeled |
| with biotinylated nucleotides. Each labeled DNA sample was hybridized |
| to a chip in hybridization buffer overnight at 50C. The hybridized |
| chips were washed several times in neutral buffer at RT, then |
| incubated with streptavidin for 15 min at RT. Following two washes at |
| 35C in neutral buffer, chips were incubated with biotinylated |
| anti-streptavidin antibody for 15 min at RT, followed by another two |
| washes at 35C in neutral buffer. Then, chips were stained with |
| streptavidin-Cy-chrome conjugate for 15 min at RT, followed by two |
| washes with neutral buffer at 35C. Chips were incubated 30 min at 37C |
| in low-salt buffer, followed by a wash with neutral buffer at RT. |
| Hybridization of the sample to the array was detected using a confocal |
| laser scanner. |
| Individual genotypes for a SNP were determined by clustering |
| measurements from multiple scans in the two-dimensional space defined |
| by background-adjusted intensities of the perfect-match features for |
| the reference and alternate alleles. We used a K-means algorithm to |
| assign measurements to clusters representing distinct diploid |
| genotypes. We determined an optimal background value for each SNP |
| that minimized the variance within the assigned genotype clusters. The |
| K-means and background optimization steps were iterated until cluster |
| membership and background estimates converged. To determine the |
| appropriate number of genotype clusters, we repeated the analysis for |
| 1, 2, and 3 clusters, and selected the most likely solution, |
| considering likelihoods of the data and the cluster parameters. The |
| data likelihood was determined using a normal mixture model for the |
| distribution of intensities around the cluster means. The model |
| likelihood was calculated using a prior distribution of expected |
| cluster positions. |