 So, the next presentation is going to be from Lucia Hindorf talking about a new resource that has been generated and thought you would be interested in. Great. Thank you, Mark. Can everyone hear me? Okay. Fantastic. So, good morning. I was asked to give a brief presentation on a new resource that we thought would be of interest to council that stems from a joint NCBI NHGRI collaboration. Before I get too far into my talk, I just want to see if Doug Hoffman from NCBI was able to join us on the phone. Doug, are you there? Okay. I don't hear him. He was going to try and join us. Maybe he'll be able to join by the end of the presentation. We were hoping he'd be available for questions. Okay. I'll go ahead and get started. Okay. So, the rationale for developing this resource started with a deluge of findings from genome-wide association studies. So, as you all know, the GWAS design has been a powerful discovery engine culminating in over 4,300 variants discovered that have been associated with over 200 human traits and diseases. Despite this, the functional implications are rarely clear. And thus, replication, fine mapping, and functional studies are crucial next steps to evaluate biologic plausibility, for example, to prioritize which signals to follow up, and ultimately to assess how genome acknowledge fits in the context of improving human and public health. And what we hope to do was to integrate GWAS data with other genomic data to help build this knowledge base. Therefore, we added a layer of genotype phenotype association data from both our GWAS catalog as well as DBGAP from NCBI to existing genomic NCBI resources. And this is just kind of one part of the knowledge base that's critical to applying genomics to public health. So, the goal of FIGENI was to develop a user-friendly integrated online resource that would be of use to population scientists and clinicians who produce or use GWAS results. These people might not necessarily be developing their own informatics databases but would be informed consumers of GWAS results. And here I'm showing you the front page of the FIGENI resource with the search interface. And here's the URL. As you can tell, we wanted to keep it simple and user-friendly. And thus, users can search on basically two properties, either the phenotype side of things on this side or the genotype side of things on this side. I'll show you the search interface in more detail shortly. So here are the databases that comprise the FIGENI resource as well as their record counts, their current record counts. The phenotype searches are linked to the association results from both the catalog, the GWAS catalog, and DBGAP, as I mentioned before. And the data on genomic variation on genes and on mRNA expression are integrated from DBSNP, entree gene, and the GTX resource, accordingly. So here I'm showing a distribution of the SNP association results and how they fall in various broad phenotype categories just to give you a sense of where the data are at. As you can see here from the pie chart on the left where I've labeled all of the categories with a 4% frequency or greater, most of the records are concentrated in just a few categories, including chemicals and drugs and diseases related to the digestive system, the eye diseases, immune system, mental disorders, nervous system disorders, cancers, as well as a skin and connective tissue disorders. So I also wanted to point out here that some associations can belong to multiple trait categories and I've weighted them appropriately here. So again, as you can see that most of the association results, at least, that we filter on are belonging to just a subset of the phenotypic categories and that obviously reflects what we see in our GWAS catalog as well as what's submitted to DBGAP. Okay, so now I'd like to give you a more detailed look at how Fee Genie works. I mentioned that you can search on phenotype, both the broad and narrow categories, as well as genotypic properties such as the actual gene, the SNP, the RS number, or the chromosomal range. You can further more filter the association results on the p-value of the association and the SNPs on functional class if you want to more efficiently search for high priority results. We've made the data tables downloadable so that people can explore the results further and included hyperlinks to the record and source databases from NCBI in order to help the user browse through various properties. We wanted to make the user interface simple as I mentioned and we felt that a portal interface would accomplish that and I'll show you what that looks like on the next slide. And then finally, NCBI has an interactive sequence viewer that allows the user to explore the genomic context of a particular genomic region in more detail and we wanted to incorporate that as well into Fee Genie. So here's a sample search on Alzheimer's disease where I've just selected from the drop-down nervous system diseases in Alzheimer's disease. As I mentioned, this is a portal interface. There's actually a whole series of little boxes that span down the page, but I can only show you a couple at a time. So at the top, you see the search summary which shows you again what you searched on and then a summary of the various results that are returned. So there are 5, sorry, 122 association results. And of the first page of those results, 55 genes are returned, 45 SNPs, and then two EQTL data results. I want to point out to you a few properties of the portal interface. So you can, for example, move the different sections up and down. There are information links that will tell you more information about the particular section that you're in. So this is kind of our documentation. You can download data tables, as I mentioned, and to CSV file formats to manipulate the results further. And everything you see in blue here is a hyperlink. So this is a link to DBSNP, Entree Gene, the Genomic Browser, an association browser that DBGAP has, the source database, which in this case is the NHGRI GWAS catalog, as well as the PubMed abstract of the studies published. Further on down the page is the genes table, which gives you properties related to the gene, the 55 genes that were returned, including their location, and then the OMIM number were available so that you can find out more about the context of that gene. Further on down the page is this ideogram, which shows you how the 55 genes are distributed among the different chromosomes. This is interactive, so if you select one of these little purple triangles, it will bring up this sequence viewer. So this is a customizable view. You can configure the different tracks that are represented, but as a default, what you see here are the different SNPs that are in this genomic region, as well as cited variants mentioned in PubMed, the association results from either DBGAP or the GWAS catalog, and where available they've included recombination rates from face to hat-mount populations, as well as kind of the genomic and mRNA and protein transcripts down here, and these are all interactive, so you can click on them and see the actual accession sequence, for example. You can zoom in and out of this region, and as I mentioned, you can configure tracks and add your own, so forth. So further on down the page, you see properties of the SNPs that are returned by Fijini, including, let's see, the functional class of the SNP, as well as whether the SNP has been validated in 1000 genomes or hat-map. If you see a blue link here in the genotypes column, that will take you directly to genotype frequencies that have been reported to DBSNP, so for example, 4,000 genomes or hat-map or any other user-submitted frequencies. And then lastly, here are the two EQTL data results that are returned from GTECs, and as you see here, two SNPs are associated with expression in lymphoblastoid tissue, and then you can see the links out to GTECs here and here. So all in all, this has been a very fruitful collaboration for both NCBI and NHGRI, with NCBI doing the heavy lifting in terms of implementation. We plan to continue regular meetings to discuss improvements to this resource, as well as drafting a manuscript to let the scientific community at large know that this resource is here. So we do plan on some future improvements, as I alluded to. There is going to be a YouTube demonstration of this resource, so if you don't want to read the manual, you can just watch a quick video to see how somebody would walk through this resource. We want to incorporate additional data types and databases into Fijini to make it more integrated and user-friendly. We're working on making the ideogram downloadable so that people can customize it and use it for talks and presentations or papers. And then finally, we want to improve our documentation. So we hope that this resource will facilitate follow-up of GWAS results by the scientific community at large. I hope you'll get a chance to check it out. I couldn't finish this talk without acknowledging the crucial participation of several individuals from the Office of Population Genomics, Erin Ramos and Heather Junkins. And at NCBI, Doug Hoffman is the primary developer in Mike Fiolo's group, and contributions from Donna Maglott, Lawn Fan, Justin Pascal, and Steve Sherry, who lead the respective genomic databases. If you have any questions of a more technical nature, you're welcome to write to the help desk link at the bottom of the Fijini page that goes straight to NCBI, or you're welcome to email me and my contact information is here. So I'd be happy to take any questions. I think if Doug, were you able to join us? Oh, we still don't have Doug. I'm happy to pass any questions. I'm here. Oh, you're here, great, welcome. Yeah. So Doug's here to answer any technical questions you might have as well. Yes, Mike. Yeah, I don't have a question, just a comment. I think this is really cool. Thanks very much for doing it. Oh, great. Thank you. Thank you very much.