 Hello, my name is Tim Griffin. This is my talk on using Galaxy and a multi-omic workflow for identifying neoantigen peptides for immun oncology. The work you will see here is a team effort at the University of Minnesota. Primarily, David Larkin-Spada is the PI on the funded grant that is supporting most of this. Others who have contributed are part of the Galaxy for Proteomics team or the Galaxy P team at the University of Minnesota. So the work is focused on these malignant peripheral nerve sheath tumors, MPNSTs, and these develop through a loss of this NF1 allele, this gene, and it follows this development progress where it goes from benign to pre-malignant and in some percentage of these to a fully malignant tumor phenotype. So the idea of this work is to develop some treatments that can intervene on this development of these types of tumors. The hypothesis is that there are non-normal epigenetic alterations that ultimately produce non-normal mRNAs that will then be translated into peptides that are presented as neoantigens to the immune system. So how does this work? A little bit about behind this is that due to some genetic alterations, we see these epigenetic changes that lead to altered chromatin state and loss of regulation and ultimately the expression of these non-normal, non-normally processed RNA sequences. And this picture sort of shows the overall concept here. So within these patients, due to these alterations, we have an expressed RNA and as a ribosome goes down that RNA, it's making protein. It will hit some of these neojunctions. These are generally splice variants that cause the reading frame to go out of frame. So these frame shifts that start to then make these very unique amino acid sequences in the proteins. So it's also known that within these types of systems, the treatment by this antibiotic, Gentamycin, will cause the ribosome to actually go beyond these premature stops and actually lengthen out the amount of non-normal or variant sequence amino acid sequence within the proteins that are being made. We call these cryptic neoantigens. So the basis of this work is to really leverage that idea and use a multi-omic approach to ultimately come up with a treatment for these types of tumors. So to use RNA-seq in a mouse model to start to identify these non-normal RNA sequences and predict potentially these expressed neoantigens, but also very importantly, confirm these neoantigens peptides are being expressed by using mass spectrometry in the same system and identifying peptides that are bound to the HLA or the MHC class 1 complex that are presented to the immune system. We can identify and verify some of those peptides. We can then take some of those, determine whether they elicit an immune response, and ultimately, potentially come up with a prophylactic vaccine. So a panel of these neoantigens peptides that could be used to intervene as a treatment in these cancers. To do this, we've been working on a protocol and using some protocols in the literature to enrich these MHC, also called HLA-1 complexes of proteins that bind to these neoantigen peptides that are presented to the immune system. So that's what this is showing, that if you enrich using antibodies, the MHC complex, you can then elude off the bound peptides, put those into the mass spectrometer and use tandem mass spectrometry to try to identify those bound peptides. And that's where the bioinformatics comes into play. So we have developed a workflow that's galaxy-based where we take the RNA-seq data, we take the accompanying peptide level mass spectrometry data, these enriched peptides, we run them through two different workflows. One takes the RNA sequencing reads and uses various tools in Galaxy to predict those non-normal peptides that potentially are expressed that would be bound to the MHC complex. We then use a tool called PEP query to take these candidates and test whether or not the MSMS data, the peptide data, shows any evidence for these being present in the samples. So that's a candidate-based identification. We also developed a more open-ended discovery-based identification where non-normal transcripts, all of them are translated out into all the potential non-normal peptide sequences. We then take the mass spectrometry data and try to match these, so we eliminate this prediction step here and keep this a little more open-ended and try to discover what are all the different variant sequences that are bound to that HLA1 complex. We've been applying these bioinformatics tools to some literature, some data that was already published on neoantigens. This is just a really little small representative taste of that data. These are the neoantigens sequences, and as you can see, they differ quite a bit from the predicted reference sequences. So this is what the type of data we hope to get out of our actual tumor data. So with that, we are continuing this work. We're using RNA-seq level data to predict these peptide neoantigens, and in the midst of testing this enrichment step for the peptides in order to do the galaxy-based analysis. Thank you.