Figure - PMC

Skip to main content

An official website of the United States government

Here's how you know

Here's how you know

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

View full-text article in PMC

. 2023 Dec 9;39(12):btad743. doi: 10.1093/bioinformatics/btad743

Search in PMC
Search in PubMed
View in NLM Catalog
Add to search

© The Author(s) 2023. Published by Oxford University Press.

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.

PMC Copyright notice

Figure 3. — (a) Violin plot of AUROC scores grouped by minimum edit distance to train dataset. The large (red) dots are (unweighted) averages of the scores for the given minimum edit distance. (b) Comparison to TCRconv. TCRconv was trained on three subsets of 30 epitopes from the $D_{α β, β}$ dataset and compared to EPIC-TRACE trained on full $D_{α β, β}$ folds either with all or reduced input features. The y-axis shows average per epitope AUROC values of frequency binned epitopes with standard error. (c) Comparison of models trained with all datapoints or by discarding epitopes with <15 TCRs from training for TPP3. Models were trained with or without MHC information. The y-axis shows the average per epitope AUROC with standard error.