Neodb

Neodb integrates a curated database and computational models for neoantigen discovery and immunogenicity prediction to support cancer immunotherapy research.

Key Features:

• Extensive database of experimentally validated neoantigens: A curated collection compiled from comprehensive literature searches of experimentally validated neoantigens linked to somatic DNA alterations in tumors.
• Immuno-GNN model for immunogenicity prediction: Immuno-GNN employs a graph neural network with an attention mechanism to model spatial interactions between human leukocyte antigen and antigenic peptides for neoantigen immunogenicity prediction, representing the first reported application of GNNs in this context and reported to outperform existing methods.
• Integrated Tools module for neoantigen prediction: A Tools module consolidates commonly used computational tools for neoantigen prediction.
• Driver-Neo module: Driver-Neo provides a curated collection of public neoantigens derived from recurrent driver mutations and includes filtering features to identify candidate neoantigens.

Scientific Applications:

• Neoantigen immunogenicity studies: Benchmarking and analysis of neoantigen immunogenicity using experimentally validated data and predictive models.
• Neoantigen-based cancer immunotherapy development: Prioritization and selection of candidate neoantigens for personalized cancer vaccines and adoptive T cell therapies.
• Translational research: Integration of mutation data and validated neoantigens to inform clinical and preclinical studies.

Methodology:

The database was constructed through comprehensive literature searches and integration of commonly used neoantigen discovery tools; the Immuno-GNN was developed as a graph neural network with an attention mechanism modeling spatial interactions between human leukocyte antigen and antigenic peptides.