Targeted Immunotherapy is entirely dependent on the identification of suitable TAAs or even better, of TSAs, also referred to as neoantigens. This is currently an extremely challenging task, mainly based on using experimental data from DNA and RNA sequencing technologies to generate large lists of neoantigen candidates, followed by heavy bioinformatics to predict how well each neoantigen candidate would progress through the MHC processing and presentation pathway. Up till now, most of the candidates generated in this way fail to elicit any relevant immunogenicity. The vast majority of candidates likely fail because, being based on single point mutations (as most of them are with the current state-of-the-art), and thus single amino acid variants (SAAVs), they are not that different from the respective non-mutated self-antigens tolerated by the immune system. Thus, very likely there is no T cell clone available capable of recognizing the putative neoantigen. Other problems could be due to intricacies of the MHC processing and presentation pathway, such as absence of a suitable HLA type able to present the putative neoantigen.
HiRIEF LC-MS, RNA-seq, DNA-seq (+ gene panel seq), bioinformatics pipeline (python and R).
We used our proteogenomics-based pipeline to find potential neoantigens in a cohort of 141 NSCLC patients. Identified putative neoantigens were next in silico validated in 3 public NSCLC proteomic datasets and filtered based on 30 normal-tissue proteomic datasets to exclude unannotated normal peptides. After MHC-I binding prediction, our analysis revealed a list of high-confidence neoantigen candidates which will be subsequently validated in vitro and in vivo.
We identified a list of high-confidence non-canonical peptides in the cohort of 141 NSCLC patients which will be further validated to choose most promising candidates to advance into clinic.
Lehtio group (ONKPAT, Karolinska Institutet).
Karolinska Institutet, Scilifelab.
The author has declared no conflicts of interest.