 Hi everyone. So today, I'm going to present you some of the results from the data and the analysis that Harvard GCC has done for various tumor types. So I'll start with an overview of the data that we have generated and our pipeline. So so far, we have sequenced 14 subsets of 14 TCJ tumor types. We have performed paradigm called genome sequencing on false coverage of 6 to 8. And this was done not only to this end, but primarily to detect structure variants. So for analysis, we used two tools, Break Dancer, which detects discordant read pairs, and also Meerkat, which gives the advantage of detecting the split read that spans the junction break point. So for this analysis, I included data on 12 tumor types listed here. In total, there were 1,200 tumors. And they're much normal samples, either blood or tissue that were used. In the majority of these tumor types, we had at least 100 samples in each case, and for three around 50 sample pairs. So here's an overview of the distribution of all the sequence samples regarding the number of structure variants that were detected. In total, we detected almost 37,000 structure arrangements out of which something like 12,000 events were zinzine, meaning they had both break points within zine region. So the most rearranged tumors were the gastroesophageal, bladder, and melanoma. And the more quiet, as expected, was the thyroid. So we tried to test the specificity of our pipeline. And we used different approaches. For example, we gathered the events that were called by both breakdowns of Meerkat. So we could have more confidence on those, or events that they could also be cross-confirmed with RNA-seq data. But ultimately, we did also lots of bench work with PCR amplifying of the adduction fragment with patient-specific primers. So we were able to cross-confirm or validate something like almost 8% of the events that we detected. So then, for this analysis, I included data on rearrangements that have to do with the RAS pathway. And I chose to do so because of the significance of this pathway. We know that there are so many tumor types that have RAS aberrations, some to a high extent, like pancreas cancer. But practically, almost in all TCGA working groups, we hear about RAS rough aberrations that activate the RAS map kinase pathway in these solid tumors. So we wanted to examine cases that still have RAS activation, but other RAS DNA changes are not obvious. And we focused on these 12 tumor types that I've shown you. And we've seen that almost 200 samples of the cases of what we have sequenced had at least one somatic gene-zine RAS rearrangement. So then the question is, what was the relationship of this RAS rearrangement that we see in comparison to other RAS aberrations? But before coming to that, just to how we defined the RAS pathway, so we heard yesterday about the needs to compile all the given established knowledge on certain cancer pathways, till this goal has been reached. We are using the RAS pathway that NCI's RAS initiative has put together. And in red, there are the genes that comprise structure arrangements in our data analysis. So that includes not only RTKs that draw most of the attention, but also genes in the source RAS RAS axis, and also genes upstream of RAS that either inhibit, like NF1, or activate RAS. So we then compared these cases with RAS rearrangement to cases with other RAS DNA aberrations. Sorry. And before coming to that, I'll just show you some examples of the events that we are seeing. So here, you can see some cases of genes that are well known, like BRAF or RAF1, with structure rearrangement, but also other interesting cases upstream of RAS, including disruption of tumor suppressor genes like NF1. So with this panel, you can see with arrows the breakpoints in its gene. And then in red and blue color, expression data as this course. So now I'm coming to the question whether the RAS rearrangements co-cure with other RAS amatic aberrations. So this subject has already been touched by another speaker yesterday saying that it seems that aberrations within one single cancer pathway seems to be mutually exclusive across different cases. And that's what we also observe here if we compare the RAS structure rearrangement with the RAS RAF mutation simplification. So we see that the RAS rearrangements are mutually exclusive to most of these somatic aberrations. We also heard about how structural variations can affect in different ways transcription. Also, you some examples with structure rearrangements that affect practically the 3GTR in some well-known oncogenes like FTFR3 and FTFR2. So here in this slide, I've got an example of a translocation in a low, great glioma patient where there is a fusion of FTFR3 with LF3 gene. So the interesting part and the beauty of the DNA-based analysis is that you can actually see the original change that occurred in the genome and that resulted later in the fusion transcript that can be oncogenic. So on the RNA level, we see the fusion of FTFR3 to action 2 of LF3 gene. But the actual genomic breakpoint lies within the 3GTR of FTFR3. And due to alternative splicing, the final fusion transcript is missing its 3GTR. So now there is already experimental data for other gene fusions like FTFR3, TAC3, and how these are oncogenic because of the loss of microRNA control over FTFR3. So that could be, assumingly, a similar way through which this rearrangement and fusion are all containing. In a similar way, we detected an inversion in a gastric case where FTFR2 is linked to WDR11 gene. So in this case, the genomic breakpoint is in the last action of FTFR2. And interestingly, there is a new stop codon within action 17. And finally, FTFR2 is losing its 3GTR and acquires a new 3GTR, practically. Part of which is intran 17. And part of which is a fragment of the 3GTR of WDR11. So, assumingly, this FTFR2 is also liking control of microRNAs on its original 3GTR. However, such cases would require experimental validation to show that there are a congenic. And we generally felt this was very important for all the events that we detected. So that's why we initiated a collaboration to try to functionally validate some of the raster arrangements that we detected. And for that, we're currently collaborating with Henry Liu and Kenny Scott from Baylor Institute. And that group provides a very interesting and robust experimental pipeline that can validate mostly rearrangements that result into a fusion transcript that is oncogenic and that drives cell proliferation. So this system takes advantage of a large set of gateway of CDNA clones, from which it is possible to PCR amplify the part of the gene that is needed. And then through a multi-site LR recombination, the final fusion gene can be cloned into this nation vector and finally the expression vector. So these expression vectors are being transfected to BIF3 cell line, a mouse line that is normally dependent on interleukin 3. So once the cells are deprived from interleukin 3, then the transformed cells can be selected. And the oncogenic protein, fusion protein, can drive cell proliferation, which can be measured by an luminescent cell viability as seen. So some of our raster arrangements are already part of this pipeline. So some results for the track 1 rough 1 rearrangement, which occurs through an inversion in chromosome 3 in a melanoma patient. So both in the whole genome sequencing data and in the RNA sec, we can see both products of this rearrangement, meaning track 1 rough 1 and its reciprocal partner, rough 1, track 1. However, only the track 1 rough 1, which after all contains the kinase domain, scores higher in the cell viability assi, comparable to the BCR-ABL fusion and not the rough 1, track 1. So the track 1 rough 1 can promote cell proliferation in our cell system. And furthermore, focusing on the RAS pathway in its activation, it does activate RAS pathway. And here, I'm showing you the increase of levels of phosphorylated air in the transformed cells. Henry and Kenny also performed an assessment about drug sensitivity of the transformed cells. And they were able to show that track 1 rough 1 cells do so are sensitive to the making heaps or traumatinomy. So to summarize, RAS mutations are found generally at high frequency in many different cancers. And we just proposed that structural arrangements also are a mechanism that alternatively can activate the pathway, so in order to have the full view of the RAS activation across different tumors, we could take them into consideration. We've shown that RAS arrangements are largely mutually exclusive to other RAS aberrations. And we're already putting our data in functional studies to show which events are functional and which events can be prioritized. So this is a need for all data that come out from studies like the TCGA. This need will be growing in the future. They need to functionally show that our high quality data have impact on the patients. So you can join me also on poster number 59 with more questions. Thank you. I have just a comment to make. I really would advise some caution towards the thinking that the primary function of this fusion is simply to eliminate the three prime regulatory regions. And there are several reasons for this, certainly the most important one being that the simple over expression of the RTK gene in the case of FGFR is of a much less value in terms of oncogenic role than the expression of the fusion. What we typically see is that, in fact, is the fusion gene, which is extremely transforming the fusion protein and not the expression of the individual partner. On the other hand, the three prime gene is clearly very important in providing multiple oncogenic activities such as homodimerization, as well as aberrant activation of downstream signaling. I mean, the idea that there is a simple activation of the canonical RTK pathway here, most of the times, I should say, is not correct. So these fusions clearly have novel functions most of the times, and the elimination of regulatory regions is probably just an incident of the genetic rearrangement. Thank you for this input. Unfortunately, the functional system that I've shown is not appropriate to test this. However, I think we should test more about what's the functional impact and on which cancer phenotype they're actually contributing. Thank you, Angeliki. Our next speaker is Julia Shee. She is going to talk about somatic copy number alteration in UVL melanoma.