 All right, after that spectacular start, we're happy to be able to introduce GeneSpot, which is a web-based portal for interactive exploration of TCGA data from a gene-centric point of view. I'm Brady, and I'm joined by Hector. I'm going to describe the motivation for this application, as well as the key features behind it. And then Hector will be able to go over a demo, hopefully, if it comes together. And if not, then show some screenshots and discuss the implementation that we've taken for this application. All right, so there are many cancer biologists out there who want to take their favorite gene and query what the data are in the TCGA for that given gene. And some example questions could very well be, across all the tumor types, what does the mutation profile look like? Are there significant copy number aberrations? What are the data drive statistical associations? What would plots of just the raw data look like? Can they be downloaded? And so on. There's a lot of very gene-specific questions that can be asked across all these data sets. Part of the issue at the moment is that many of the data repositories are organized in a very tumor-specific or sample-centric point of view, and so we're just trying to take an orthogonal view from those and provide a new way of looking at the data. Among those lines, a typical workflow to look at a gene-centric point for TCGA would be to download the data from a repository, being DCC or road fire hose, for example, and then parse and process all those data, and finally put together all the different features, being clinical, molecular, whatever, across all the different samples and all the different tumors. And then from that, be able to extract out the gene of interest and then do the analyses and create the plots and move forward from there. Part of the challenge is that it's actually not very trivial to do this. The data sizes aren't that large, but the time to answer these types of questions can take quite a while, and so estimating 10,000 samples and about 500,000 data points for each of those then gives you 5 billion data points, most of which aren't relevant for whatever gene you're interested in evaluating. So the approach we've taken is to create an interactive web portal, which may or may not work given the wireless, we'll see. And then from that, you can query a certain gene or sets of genes, and then start to explore the data from there without the need to download or install software. The other part of the approach that we've taken is to provide basically a workspace or a controllable canvas where you can select certain types of views given the gene that you're interested in and what the data are telling you and then move those, remove those, expand, minimize, whatever you would like, and from those create sessions that you can then come back to at a later time or share and collaborate with others to show what's going on. The other aspect, too, is to consider data direct access or provenance where you want to see what are the input files, what are the tab delimited data look like, how can I then do my own subsequent analysis from those data, but querying them in a very accessible way. A few example screenshots that will lead into the demo that Hector will have is not so much to focus on the biology, but just to give some quick snapshots of what you can look for. And many of these views are inspired by others within TCGA community or typical views that you would see when you're looking at any given gene across the different tumor types. So here we have FBXW7 and can look at what its mutation profile is and have also in the context the different protein domains on the bottom and see where the mutations are clustering, what types of mutations there are. And the web portal you can hover over many of these points and see what the details would be, sample identifiers and whatnot. Also integrating in with other sources like Mutesig rankings where you can query and say here's my favorite gene, what are the different ranks and the different tumor types that can help prioritize subsequent analyses that you might have to then drill down on and spend more time in the data portals downloading data and whatnot. Additionally, seeing the context of your gene within all the others within rankings, for example, Mutesig top 20, which is taking a lot of the different tables that have been created, but looking at them in a gene centric point of view. Also copy number, primarily using just a coming from broad analyses. And looking at which tumor types they're significantly amplified or deleted in, as well as inspired by the Anko prints, looking at a per sample view. And again, this is just to show an example of how you can take different sets of genes or individual genes and look at what their profiles are across different tumor types throughout all the data. Let me switch to the other laptop real quick. It's going to be interesting. Change the zoom level a little bit just for the presentation. So as you can see, can you see that okay? Yeah. So we're trying.