Table 2.
Peptide | McPAS | Peptide | VDJdb | ||||
---|---|---|---|---|---|---|---|
AE | LSTM | ERGO-I | AE | LSTM | ERGO-I | ||
LPRRSGAAGA | 0.827 | 0.820 | 0.772 | KLGGALQAK | 0.765 | 0.758 | 0.731 |
GILGFVFTL | 0.876 | 0.886 | 0.843 | GILGFVFTL | 0.874 | 0.864 | 0.820 |
NLVPMVATV | 0.884 | 0.886 | 0.835 | NLVPMVATV | 0.818 | 0.827 | 0.686 |
GLCTLVAML | 0.861 | 0.871 | 0.816 | AVFDRKSDAK | 0.750 | 0.737 | 0.695 |
SSYRRPVGI | 0.975 | 0.975 | 0.980 | RAKFKQLL | 0.820 | 0.774 | 0.828 |
The results are for AE-based and LSTM-based models, including TCRα, V, J genes, MHC and T-Cell type data usage. A comparison to ERGO-I best result is shown. Bolded values are the best results on a certain peptide.