NeoFox: annotating neoantigen candidates with neoantigen features

Lang, Franziska; Riesgo-Ferreiro, Pablo; Löwer, Martin; Şahin, Uğur; Schrörs, Barbara

doi:10.1093/bioinformatics/btab344

Cited by 15 publications

(24 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Neoantigen candidates from SNVs and INDELs were identified with an in-house property pipeline 25 (“iCaM2”) and candidates from fusion genes were identified with EasyFuse 29 . Neoantigen candidates were annotated with features and subjected to prioritization algorithms using NeoFox 15 .…”

Section: Resultsmentioning

confidence: 99%

“…Our results indicated that integrating the quantitative approach relying on the neoantigen candidate load and the qualitative multiple instance approach may improve the prediction of the response to ICB. Furthermore, a limited set of neoantigen features was integrated into the model approach in this study, mostly targeting the linear sequence of neoantigen candidates and rather focusing on the interaction with MHC molecules 15 . Integrating clonality information 11 , structural features 45 and future features that specifically model the interaction between the MHC bound neoepitope candidate with the TCR repertoire may improve predictions.…”

Section: Discussionmentioning

confidence: 99%

“…The harmonic mean of best predicted MHC-II binding rank across the MHC-II genotype with an in-house property pipeline 25 ("iCaM2") and candidates from fusion genes were identified with EasyFuse 29 . Neoantigen candidates were annotated with features and subjected to prioritization algorithms using NeoFox 15 . neoantigen candidates ("-") or only neoantigen candidates confirmed in the RNA-seq ("+") were used.…”

Section: Phbr_iimentioning

confidence: 99%

“…Neoantigen candidates from all mutation types were annotated with published neoantigen features and prioritization algorithms using the NeoFox (v0.5.3) toolbox 15 . The predicted neoantigen candidates, MHC-I and -II genotypes of the patient and the tumor type were provided as input.…”

Section: Annotation Of Neoantigen Candidatesmentioning

confidence: 99%

See 3 more Smart Citations

Multiple instance learning to predict immune checkpoint blockade efficacy using neoantigen candidates

Lang

Sorn

Schrörs

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

A successful response to immune checkpoint blockade treatment (ICB) depends on the functional re-invigoration of neoantigen-specific T cells and their anti-tumoral activity. Previous studies showed that the patient's neoantigen candidate load is an imperfect predictor of the response to ICB. Further studies provided evidence that the overall response to ICB is also affected by the qualitative properties of a few or even single candidates, limiting the predictive power based on candidate quantity alone. To our knowledge, this is the first study to predict the response to ICB therapy based on qualitative neoantigen candidate profiles in the context of the mutation type, using a multiple instance learning approach. Multiple instance learning is a special branch of machine learning which classifies labelled bags that are formed by a set of unlabeled instances. The multiple instance learning approach performed systematically better than random guessing and was independent of the neoantigen candidate load. Qualitative modeling performed better in comparison to the quantitative approach, in particular for modelling low-abundant fusion genes. Our findings suggest that multiple instance learning is an appropriate method to predict immunotherapy efficacy based on qualitative neoantigen candidate profiles without relying on direct T-cell response information and provide a foundation for future developments in the field.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Phbr_iimentioning

confidence: 99%

Section: Annotation Of Neoantigen Candidatesmentioning

confidence: 99%

See 2 more Smart Citations

Multiple instance learning to predict immune checkpoint blockade efficacy using neoantigen candidates

Lang

Sorn

Schrörs

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…MHC binding is performed with the tool netMHCpan 4.1 [16] with the following parameters 'netMHCpan -p {fasta} -a {hla_allele} -s' for every epitope of 8, 9, 10 and 11 amino acids using NeoFox [20] API. We keep only the rank of the best binding for every allele independently.…”

Section: Methodsmentioning

confidence: 99%

In silico analysis of predicted differential MHC binding and CD8+ T-cell immune escape of SARS-CoV-2 B.1.1.529 variant mutant epitopes

Riesgo-Ferreiro

Gudimella

Bukur

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Introduction: The B.1.1.529 (Omicron) SARS-CoV-2 variant has raised global concerns due to its high number of mutations and its rapid spread. It is of major importance to understand the impact of this variant on the acquired and induced immunity. Several preliminary studies have re-ported the impact of antibody binding and to this date, there are few studies on Omicron CD8+ T-cell immune escape. Methods: We first assessed the impact of Omicron and B.1.617.2 (Delta) variant mutations on the SARS-CoV-2 spike epitopes submitted to the Immune Epitope Database (IEDB) with positive outcome on MHC ligand or T-cell assays (n=411). From those epitopes modified by a mutation, we found the corresponding homologous epitopes in Omicron and Delta. We then ran the netMHCpan computational MHC binding prediction on the pairs of IEDB epitopes and matching homologous epitopes over top 5 MHC I alleles on some selected populations. Lastly, we applied a Fisher test to find mutations enriched for homologous epitopes with decreased predicted binding affinity. Results: We found 31 and 78 IEDB epitopes modified by Delta and Omicron mutations, respectively. The IEDB spike protein epitopes redundantly cover the protein sequence. The WT pMHC with a strong predicted binding tend to have homologous mutated pMHC with decreased binding. A similar trend is observed in Delta over all HLA genes, while in Omicron only for HLA-B and HLA-C. Finally, we obtained one and seven mutations enriched for homologous mutated pMHC with decreased MHC binding affinity in Delta and Omicron, respectively. Three of the Omicron mutations, VYY143-145del, K417N and Y505H, are replacing an aromatic or large amino acid, which are reported to be enriched in immunogenic epitopes. K417N is common with Beta variants, while Y505H and VYY143-145del are novel Omicron mutations. Conclusion: In summary, pMHC with Delta and Omicron mutations show decreased MHC binding affinity, which results in a trend specific to SARS-CoV-2 variants. Such epitopes may decrease overall presentation on different HLA alleles suggesting evasion from CD8+ T-cell responses in specific HLA alleles. However, our results show B.1.1.529 (Omicron) will not totally evade the immune system through a CD8+ immune escape mechanism. Yet, we identified mutations in B.1.1.529 (Omicron) introducing amino acids associated with increased immunogenicity.

show abstract

Deep Learning and Transformers in MHC-Peptide Binding and Presentation Towards Personalized Vaccines in Cancer Immunology: A Brief Review

Machaca,

Goyzueta,

Cruz

et al. 2023

Lecture Notes in Networks and Systems

View full text Add to dashboard Cite

NeoFox: annotating neoantigen candidates with neoantigen features

Cited by 15 publications

References 9 publications

Multiple instance learning to predict immune checkpoint blockade efficacy using neoantigen candidates

Multiple instance learning to predict immune checkpoint blockade efficacy using neoantigen candidates

In silico analysis of predicted differential MHC binding and CD8+ T-cell immune escape of SARS-CoV-2 B.1.1.529 variant mutant epitopes

Deep Learning and Transformers in MHC-Peptide Binding and Presentation Towards Personalized Vaccines in Cancer Immunology: A Brief Review

Contact Info

Product

Resources

About