vaccines and immunoinformatics approaches

BenchMHC reproduces and evaluates pMHC presentation models using public data, showing improvements in prediction accuracy and speed.

The paper develops a computational framework to minimize CD8⁺ T cell immunogenicity in designed proteins while preserving function, using a de novo luciferase as a model.

UnibindMD integrates AI and molecular dynamics to design validated peptide binders by combining multiple stages and validating predictions through dynamics simulations.

CAI-Vax designs personalized mRNA cancer vaccines by optimizing antigen sequences based on HLA compatibility, reducing predicted junctional binding while varying performance across prediction models.

The paper identifies bottlenecks in MHC-I antigen presentation by Mtb, showing that only a small fraction of predicted binders are actually presented, and develops computational models to improve prediction.

This paper benchmarks a Gaussian process active learning framework for antibody-antigen binding affinity prediction, showing significant improvements over random selection.

The paper detects endogenous MR1B protein in 293F cells using immunoaffinity enrichment and two-dimensional western blotting, confirming its presence without overexpression.

This paper identifies barriers to generalisable TCR:pMHC prediction and proposes solutions through data analysis, evaluation setup, and metric alignment.

The paper predicts HER2 epitopes and VHH paratopes from sequence alone using ESM2 and a residue prior, achieving high accuracy for binder and paratope prediction but limited epitope performance.

This paper develops an immune world model for peptide vaccines that links molecular predictions with immune dynamics through multi-agent orchestration and native-recognition constraints.

This study uses hydrogen mass repartitioning accelerated molecular dynamics to calculate peptide HLA binding free energies, identifying immunogenic neoantigens for cancer vaccines.

This paper designs and evaluates a multi-epitope vaccine candidate against Chandipura virus using computational methods, showing strong antigenicity, low allergenicity, and predicted immune response.

This paper designs a chimeric vaccine against Necator americanus using immunoinformatics, combining epitopes from three antigens with an adjuvant to elicit immune responses and ensure safety.

NeoOmni Atlas integrates proteogenomic data to catalog tumor neoantigens across 26 cancer types, identifying novel ligands and epitopes for immunotherapy.

The paper introduces MESSI, a generative diffusion model for deimmunizing proteins by designing sequences with minimal immunogenicity while preserving structure and function.

This paper presents a pipeline for multi-epitope vaccine design that integrates multiple tools to predict B-cell epitopes and validates them against structural data.

The paper quantifies hubness in T cell receptor distance spaces and evaluates local scaling as a correction, showing improvements in retrieval accuracy but no universal solution.

The paper generates a monoallelic immunopeptidome atlas for 10 prevalent canine DLA-88 alleles, identifying 32,702 peptides and revealing allele-specific binding motifs and conserved antigen-processing signatures.

This paper develops an AI-driven vaccine design framework for Mpox, yielding three multi-epitope constructs validated for structural stability, immune binding, and expression efficiency.

This paper introduces a supervised learning denoising method to identify low-quality labels in TCR specificity data, improving model reliability by filtering noisy examples.

PREpiBind evaluates ten protein representations for pMHC-II binding prediction across multiple datasets, showing varying performance depending on the task and evaluation method.

AbAffinity predicts antibody-antigen affinity using a chain-aware sequence model, achieving high correlation on benchmark datasets and maintaining performance without structural data.

This paper develops a biophysical computational framework to identify antigen escape variants and compensatory TCR sequences by integrating mutagenesis data and structural modeling.

This paper evaluates the effectiveness of general protein language models and negative data generation strategies in predicting TCR-peptide-MHC interactions, finding that they struggle to distinguish positive from negative TCRs.

The paper estimates IgG recognition of viral epitopes across populations using a PhIP-based approach, finding broad virus detection and population-specific epitope recognition patterns.

The paper reports cryo-EM and X-ray structures of an anti-MHC-I mAb/H2-Kb complex, revealing conformational epitope details and allelic specificity.

This study identifies 59 outer membrane β-barrel proteins from B. mallei and constructs a multi-epitope vaccine predicted to induce strong immune responses.

The paper introduces TCRdenoise, a method to denoise TCR-pMHC interaction data by clustering and similarity metrics, improving the quality of training data for predictive models.

SemVac uses LLMs to predict vaccine antigens by analyzing literature-derived protein descriptions, outperforming existing methods and identifying new candidates with experimental validation.