An integrated AI approach for the prediction of Clinically Relevant Neoantigens

Time: 10:00 am
day: Day Two


• Describing unique neoantigen discovery algorithms optimized for predicting neoantigen
presentation to the tumor cell surface, in an improved manner
• A machine learning ensemble approach is taken to variant calling, identifying accurately
tumor specific variants
• The AI platform is composed of numerous models integrated in deep learning layers to
predict bone fide neoantigens
• We illustrate validation in an infectious disease setting, and applications to significantly
improve the identification of neoantigen targets for personalized cancer immunotherapy