The last decade, an increasing number of cancer patients is treated with immune checkpoint inhibitors, however accurate minimal-invasive response
prediction tests are lacking. In this project, we will develop the NIPITplus strategy where we perform computational deconvolution and TCR/BCR
indexing on bulk sequencing data of blood samples to estimate the proportions of different immune-cell (sub-)types/clones. The deconvolution will
be informed by unbiased single-cell RNA profiling of the immune system. Immune cell (sub-)types/clones associated with response will be identified by
applying machine learning algorithms. It has been demonstrated that across different cancer entities common immune cell types exist that are predictive
for response, therefore, we aim to develop a multi-entity predictor of response by including patients with lung cancer, melanoma and RCC. Finally we will
monitor patients and asses the shift in immune cell proportions and TCR/BCR abundances upon occurrence of immune related adverse events during
immunotherapy.