| | Silhouette score 2 | S/N 3 | Genotype calls 4 |
|---|
| | Average | Median | Highest | Average | Highest | Correct | Errors |
|---|
| | | | n | % | | n | % | n | % | n | % |
|---|
TERMIPol | 0.72 | 0.78 | 20 | 25.3 | 4.3 | 11 | 13.9 | 2337 | 98.9 | 18 | 0.8 |
Therminator | 0.69 | 0.79 | 15 | 19.0 | 3.6 | 7 | 8.9 | 2323 | 98.3 | 32 | 1.4 |
KlenThermase | 0.74 | 0.79 | 22 | 27.8 | 8.0 | 21 | 26.6 | 2346 | 99.3 | 10 | 0.4 |
ThermoSequenase | 0.71 | 0.82 | 22 | 27.8 | 8.9 | 40 | 50.6 | 2324 | 98.3 | 34 | 1.4 |
- 1 Duplicate experiments, each with duplicate SNP assays in both DNA polarities, were performed and the results are composite values from both experiments.
- 2 The Silhouette scores were calculated as described in Figure 1. The average and the median score for all SNPs are given for each enzyme together with the number of SNP assays (n) and frequency (%) where an enzyme yielded the highest Silhouette score.
- 3 Signal to noise ratios (S/N) were calculated from each spot by dividing the fluorescence intensity values from the fluorescently labelled ddNTP/ddNTPs corresponding to a true genotype (signal) by the fluorescent intensity value from the other ddNTPs (noise). The average S/N ratios are given together with the number of SNP assays (n) and frequency (%) where an enzyme yielded the highest S/N.
- 4 Number of genotype calls (n) and call rate (%). The genotype obtained from the majority of the assays was considered to be the correct one. The percentages of the samples not accounted for in the table failed to give genotypes.
