Skip to main content
Fig. 5 | Genome Medicine

Fig. 5

From: A robust deep learning workflow to predict CD8 + T-cell epitopes

Fig. 5

Relative similarity to autoantigens and tumour-associated antigens (RSAT). A Schematic of positive and negative selection of T-cells in thymus. B Number of self-peptides, such as cancer neoepitopes, autoantigens and tumour-associated antigens, retrieved from different databases. C Distribution of Match score between pathogenic peptides and best self-peptide counterparts. Contains the Cohen’s d value showing the effect sizes differentiating positives versus negatives as well as the number of positive and negative peptides in each species. Cohen’s d values describe effective sizes, which are small (d = 0.2), medium (d = 0.5) and large (0.8). D RSAT of pathogenic peptides that have comparable self-antigen counterparts (by match score ≥ 0.6). E, F RSAT can effectively discriminate pathogenic epitopes from non-epitopes of varying lengths and species. E Distribution of RSAT for pathogenic peptides having 9aa (left) and 10aa (right) in length. F RSAT of peptides derived from different pathogens

Back to article page