Guiding the selection of presented and immunogenic neoantigens. The NEC Immune Profiler Provides an Holistic Approach to Antigen Presentation Prediction. NEC OncoImmunity AS has developed an AI engine for its clinical projects, known as the NEC Immune Profiler, that predicts from next generation sequencing data true neoantigens for personalized cancer immunotherapy and cancer immunotherapy biomarkers.
The NEC Immune Profiler takes a holistic approach to the antigen presentation prediction challenge, by integrating patent pending modules of HLA binding, processing and antigen presentation in an integrated AI system. Each module significantly outperforms competing approaches.
The key differentiator of the NEC Immune Profiler, is the accurate prediction of antigens that are naturally processed and presented to the tumor cell surface. This unique approach has been validated against clinically relevant neoantigens, increasing the probaility of clinical success for our biopharma partners.
A Unique Approach to Guide the Selection of Clinically Relevant Neoantigens
The NEC Immune Profiler predicts HLA bound peptides presented on the tumor cell surface through an integrated machine learning algorithm trained on vast amounts of mass spectrometry, binding affinity and other data. The ranked prioritized list empowers and guides the selection of peptides that are both tumor specific and immunogenic.
Improved Hit Rate for Cell Surface Presentation
The integrated machine learning approach of the NEC Immune Profiler results in a greatly improved hit rate for the precise design of truly personalized cancer vaccines. Conventional pipelines that include HLA-binding and high expression cutoff typically capture 2 out of 10 possible true hits among over ten thousand possible candidates. However, the holistic machine learning approach performed by the NEC Immune Profiler, in its present form, captures 8 out of 10 possible true hits among the thousand possible candidates.
Improves Clinical Success Rates
The NEC Immune Profiler significantly outperforms competing prediction tools on clinically validated data. Here, aggregated data from 7 clinical studies that identified validated neoantigens are studied. Interestingly candidates were initially selected because they are predicted binders by the NetMHC family of tools. All positives have been shown to kill autologous tumors or be naturally processed in transfected target cells. Data set includes 725 9mer & 10mer mutated peptides, and contains 23 positives. No expression data was provided in these studies so presentation predictions were based exclusively on inputted binding & processing predictions. The NEC Immune Profiler is an HLA-agnostic tool and can make predictions for any neoantigen HLA combination.