Back

A novel HLA Class II presentation prediction algorithm deciphers immunogenic CD4 epitopes specific to KRAS G12C

Sprague, D.; Hart, M. G.; Klein, J.; Kounlavouth, S.; Vegesna, R.; Rotunno, M.; Kraemer-Tardif, L. D.; Zhou, R.; Kemp, L.; Greeene, A. C.; Araya, J.; Mantilla, A.; Adeoye, B.; Dominguez, C.; Ferguson, A. R.; Johnson, M. L.; Davis, M. J.; Lane, M.; Palmer, C. D.; Jooss, K.; Dhanik, A.

2024-12-10 bioinformatics
10.1101/2024.12.06.627073 bioRxiv
Show abstract

Accurate prediction of peptide presentation by HLA molecules is important for generation of effective individualized cancer vaccines and immunotherapies. While presentation prediction algorithms for HLA class I have been successfully applied in the context of such therapies, improved prediction algorithms for class II are needed. EDGE-II is a novel algorithm based on a protein large language model that has a learned allele deconvolution network trained on existing and new immunopeptidomics data. It delivers state-of-the-art performance on prediction of peptide presentation by HLA class II and immunogenicity elicited by CD4+ T-cell epitopes. In a patient with a KRAS G12C positive tumor treated with a KRAS G12C targeting immunotherapy, EDGE-II identified KRAS G12C class II neoantigens that elicited clonally expanded CD4+ T cells with cytotoxic transcriptional profiles post-vaccination. EDGE-II could play an important role in the development of effective cancer immunotherapies by elucidating an enriched understanding of the immunopeptidome.

Matching journals

The top 9 journals account for 50% of the predicted probability mass.

1
Nature Communications
○ 5641 papers in training set
Top 11%
15.3%
Similar papers in this journal
2
Frontiers in Immunology
○ 638 papers in training set
Top 3%
5.2%
Similar papers in this journal
3
mAbs
○ 32 papers in training set
Top 0.1%
4.9%
Similar papers in this journal
4
Nature Biotechnology
○ 172 papers in training set
Top 0.8%
4.9%
Similar papers in this journal
5
Nature Biomedical Engineering
○ 47 papers in training set
Top 0.2%
4.4%
6
Communications Biology
○ 993 papers in training set
Top 3%
4.4%
Similar papers in this journal
7
Nature Methods
○ 385 papers in training set
Top 2%
4.4%
8
Science Immunology
● 88 papers in training set
Top 0.8%
3.6%
9
Nature Machine Intelligence
○ 70 papers in training set
Top 0.8%
3.3%
50% of probability mass above
10
Briefings in Bioinformatics
◐ 354 papers in training set
Top 3%
2.8%
11
Cell Reports Medicine
○ 153 papers in training set
Top 1%
2.7%
12
Cell Reports
○ 1498 papers in training set
Top 16%
2.1%
13
Scientific Reports
○ 3612 papers in training set
Top 45%
2.1%
14
iScience
○ 1154 papers in training set
Top 15%
1.9%
15
Cell Reports Methods
○ 165 papers in training set
Top 1%
1.7%
16
Scientific Data
○ 209 papers in training set
Top 1%
1.7%
17
Science Advances
● 1243 papers in training set
Top 20%
1.7%
18
Molecular Therapy - Methods & Clinical Development
○ 38 papers in training set
Top 0.4%
1.7%
19
eLife
● 5828 papers in training set
Top 51%
1.7%
20
JCI Insight
● 277 papers in training set
Top 5%
1.5%
21
npj Vaccines
○ 67 papers in training set
Top 0.6%
1.1%
22
Cell Systems
○ 201 papers in training set
Top 3%
1.1%
23
Leukemia
○ 42 papers in training set
Top 0.6%
1.1%
24
Journal for ImmunoTherapy of Cancer
● 75 papers in training set
Top 1%
1.1%
25
Proceedings of the National Academy of Sciences
● 2444 papers in training set
Top 34%
1.1%
26
Nucleic Acids Research
◐ 1281 papers in training set
Top 12%
1.1%
27
Molecular & Cellular Proteomics
183 papers in training set
Top 1%
1.1%
28
NAR Genomics and Bioinformatics
◐ 242 papers in training set
Top 4%
0.9%
29
PLOS ONE
● 5266 papers in training set
Top 61%
0.6%
30
npj Systems Biology and Applications
○ 125 papers in training set
Top 2%
0.6%

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.