SNP and genotyping
  1. SNP discovery is performed using eight DNA samples from the Coriell Polymorphism Discovery Resource panel. Each sample is sequenced individually for each PCR product in forward and reverse direction.
  2. Only sequences with high quality (sequence quality > 30, or < 1 error per 1,000, an estimate of the error probabilities for each base-call) are used to identify SNPs using automated computer algorithms (Polyphred).
  3. Only SNPs that are identified using both the forward and the reverse reads are considered subsequent genotyping. Potential SNPs where one sequence trace is of lower quality, are re-sequenced before being considered for genotyping.
  4. Only SNPs whose minor allele is found in at least 2 of the 16 sequenced chromosomes, are being genotyped.
  5. All samples for genotyping are stored in microtiter plates, and handled using pipetting robots for PCR and genotyping setup
  6. SNP genotyping by Invader assays is initially tested using 90 samples of the Coriell Polymorphism Discovery Resource panel. Results are analyzed using computer algorithms to assign genotypes to each sample. Samples are typed in duplicate to detect discrepancies as a measure of error rate in the individual assays. Assays that fail in analysis are redesigned. This affects less than 5% of all assays.
  7. The resulting genotypes for the first 8 samples are compared to the genotypes originally obtained by sequencing. Of all assays designed to date, no discrepancies between sequencing and genotyping have been found.
  8. Assays that pass the initial test with the Coriell samples, are subsequently typed using 500 samples from the Berkeley study cohort. For each microtiter plate, 4 Coriell samples (from the initial set of 90) are included to monitor the genotyping quality. In addition, duplicate samples are included in the panel to estimate assay accuracy. So far, only one assay has yielded a discrepant genotype call in the duplicate samples.