Helper T cells play a central role in orchestrating adaptive immune responses and have been shown in recent studies to be essential to the elicitation of therapeutic anti-tumor immune responses by cancer vaccines. In an October paper in Nature Biotechnology, Ludwig Lausanne researchers Julien Racle, Michal Bassani-Sternberg and David Gfeller reported a new and more accurate method to identify the molecular signs of cancer likely to be presented to helper T cells. To develop their model, the researchers applied mass spectrometry and determined the amino acid sequences of more than 99,000 HLA-II binding peptides eluted from cells and tissues. With this data, their novel computational tool based on machine learning, called MoDec (for motif deconvolution), accurately defined HLA-II consensus binding motifs. The results were used to train an algorithm to predict the HLA-II presentation capability of peptides from a variety of tumors and pathogens. The predictive power of their method is at least twice as good as previous techniques and is being employed in Ludwig Lausanne’s efforts to develop individualized immunotherapies for cancer.
This article appeared in the April 2020 issue of Ludwig Link. Click here to download a PDF (1 MB).