Table 5.
TCR-β sequences present at ranked-frequency higher than the threshold in both training and testing datasets, prioritised by the number of coeliac samples in which they were observed in the testing and training datasets. Both training and testing datasets include CeD patients on gluten-containing and gluten-free diets. None of these sequences were present in any control samples. = indicates sequences with equal rankings.
| Rank | Sequence | Training: Number of CeD Samples | Validation: Number of Cross-Validation Folds |
Testing: Number of CeD Samples |
|---|---|---|---|---|
| 1 | IRSTDT | 4 | 20 | 3 |
| 2 | VRFTDT | 1 | 19 | 3 |
| 3 | SFRTTDTQ | 4 | 20 | 1 |
| 4 | ASSIRATDTQY | 3 | 20 | 1 |
| 5= | IRTTDT | 2 | 20 | 1 |
| 5= | LRSTDT | 2 | 19 | 1 |
| 5= | LRATDT | 2 | 19 | 1 |
| 5= | SASDSLNTEAF | 2 | 19 | 1 |
| 5= | SLRWTDTQ | 2 | 20 | 1 |
| 10 | ASSLTVTDTQY | 1 | 19 | 1 |