Seq2Neo

Seq2Neo predicts the immunogenicity of neoepitopes derived from somatic DNA alterations in cancer to support identification of targets for personalized cancer vaccines and combination therapies.

Key Features:

• Comprehensive Neoantigen Prediction: Supports prediction of neoepitopes from point mutations, insertion-deletions (indels), and gene fusions.
• Convolutional Neural Network (CNN) Model: Employs a CNN trained to predict neoepitope immunogenicity.
• Integration with Raw Sequencing Data: Processes raw sequencing data directly to derive candidate neoantigens from genomic alterations.
• Performance Validation: Demonstrates improved prediction accuracy compared to existing tools in independent validation datasets.
• Clinical Relevance: Aids identification of neoantigens to inform personalized cancer vaccines and combination therapies involving PD-1/PD-L1 blockade and to address non-responders to checkpoint inhibitors.

Scientific Applications:

• Personalized Cancer Vaccine Design: Predicts which neoepitopes are likely to elicit immune responses for designing patient-specific vaccines.
• Combination Therapy Optimization: Identifies neoantigens that could enhance the efficacy of PD-1/PD-L1 blockade and other immunotherapies.
• Cancer Immunogenicity Research: Analyzes immunogenic potential of somatic DNA alterations, including point mutations, indels, and gene fusions, from raw sequencing data.

Methodology:

Processes raw sequencing data and uses a convolutional neural network trained on neoepitope examples to predict immunogenicity from point mutations, insertion-deletions (indels), and gene fusions.