pVACtools: A Computational Toolkit to Identify and Visualize Cancer Neoantigens

Area of Science:

• Computational biology
• Cancer immunology
• Genomics

Background:

• Neoantigen identification is vital for predicting immunotherapy response and designing personalized cancer vaccines.
• This process requires integrating genomics, proteomics, immunology, and computational methods.

Purpose of the Study:

• To present pVACtools, a computational framework for end-to-end neoantigen characterization.
• To support the identification, prediction, prioritization, and vaccine design of neoantigens.

Main Methods:

• pVACtools integrates with genomics pipelines for neoantigen identification from various genetic alterations (mutations, indels, fusions).
• It employs an ensemble of algorithms for peptide:MHC binding prediction (MHC Class I and II).
• Prioritization uses mutant allele expression, binding affinities, and clonality data, with results visualized via pVACviz.

Main Results:

• pVACtools enables comprehensive neoantigen characterization from somatic alterations.
• The framework supports the design of DNA vector-based (pVACvector) and synthetic long peptide vaccines.
• Interactive visualization aids clinical users in reviewing and selecting candidate neoantigens.

Conclusions:

• pVACtools provides a modular, integrated solution for neoantigen discovery and vaccine design.
• It facilitates the translation of genomic data into personalized cancer vaccine candidates.
• The framework supports diverse vaccine delivery strategies and downstream analyses.