 All right, so now I'm actually going to go over two different use cases that might be applicable to some of the research that you may be conducting. And then we're going to let you guys start exploring screen. We have a few exercises that you can download from the ASHG tab on screen if you want to try to work through those. Also, we're free to answer any questions that you may have applicable to your individual research questions as well. So this is the first case, and this is actually a recent publication that came out in April. And this is a group that identified new candidate genes for autism spectrum disorder through whole genome sequencing. So these are rare genetic variants that they've identified and they've identified new genes that tend to be near or containing these variants. One of interest that they reported of 18 was a myo5a. And so this is one of the myosin 5 heavy chain genes. So it belongs to this larger super family of myosin genes. And if you kind of remember back to basic biology, myosin is one of these motor proteins that's involved in transport itself. So at first glance, this doesn't seem to be necessarily a gene that is intrinsically would be linked with autism. So it's something that you may want to be exploring more to see what sort of cell types express this gene and additional information such as that. So you can actually use screen to investigate gene expression. So if we go to screen here, you can search in the search box for this gene for myo5a. You can see there's actually autocomplete options as well. So you can search for this gene in human. And when we search for this gene, it's going to return CREs that are nearby this gene as well as give you these tabs to investigate gene expression through RNA seek data and rampage data. So if you actually look at the expression of this gene, we see it's expressed in a lot of different cell and tissue types. We can change here the scale to TPM. The top one actually though is in this type of brain neuron cells. This is specifically for the GABA pathway specific neuron. However, it's also expressed in a lot of other human adult tissues. We can look at, for example, rampage data as well. So this is going to be capturing the five prime end of transcripts. So it may be capturing, for example, RNAs that may be degraded over time or can't be captured by traditional RNA seek. And here we have the data for all the different transcription start sites for this gene. We can scroll through them. We see this top one here is has really high expression in neural cells. And then some of these other transcripts, for example, are also expressed at much lower levels, but some level in brain tissue. So it seems to be there's definitely some sort of neural role for this gene, but it's a little bit difficult to see just using the human data. So what's great about screening is it also incorporates all of the mouse data. So we can search, for example, for a CRE of interest here. So this would be, for example, a promoter-like CRE. It has really high H3K4M3, and it's proximal to the gene. So we can click this CRE. We see that it has high activity once again in neural cells, but also other cell types. And we can actually go over to mouse to see if there's an orthologous mouse CRE in there is. So if we click this CRE, it will bring us down to the mouse page. And we see that it's at the promoter of the orthologous mouse gene MyO5B. This time, if we look at gene expression, we have a little bit of a different pattern. So, as Jiping had mentioned before, we have this nice matrix of embryonic mouse data. And we can see, in addition to some blood cells, we have really high expression in embryonic mouse brain tissue. So it seems that the expression of this gene, in addition to other cell types, is really high during brain development. And so what we can do is explore this further by using the differential gene expression app, which is this delta symbol here that Michael showed before. So you can actually pick different tissue types to look at the expression for this gene. So, for example, here for interested, maybe in midbrain, we have, for example, from time point 11.5 all the way to birth. So we can select E11.5, and we can go all the way, for example. We'll go halfway to 14.5. So if we scroll down here, we see that this gene MyO5A actually increases expression over brain development. And we can see it corresponds to increased activity in these promoter-like CREs here, as well as these more distal enhancer-like CREs. So even just by looking at initial human gene expression, we couldn't get the whole picture for what this gene might be doing. It seems by incorporating the mouse data as well, we can maybe try to figure out a role for this gene in autism. And more importantly, it enables you to develop hypotheses for all these results. So if you have a list of maybe 50 genes, you want to prioritize actually testing some of these in one of your model organisms, you can actually use screen to try to prioritize which ones you want to test. So that's sort of at the gene level of one use case. And this is another use case looking at a GWAS variant. So this is a bit of an older study from 2010. And this study was looking at the relationship between human genetic variants and levels of cholesterol. So this particular study we've imported into screen already. I'm going to go back to the main page here. So here we can find this study by actually just searching cholesterol. And we have a whole bunch of actually studies here that have measured different levels of cholesterol. This one here is the Teslovitch study. So what we have here are recommended cell types to investigate. And this is based on enrichment of GWAS SNPs compared to a background. And if we look here, we can see the most significant ones are all from liver. So we have the right lobe of female liver here, an adult liver from roadmap, HEPG2, which is a liver cell line. So we can click one of these and we'll give us all this CREs that overlap a SNP as well as our active in liver. And you can sort these as well. And so you can actually start to try to investigate individual SNPs that may be causal based on overlapping epigenomic data. And this is an example of one here. So this is a distal CRE with the D symbol, but it has high DNA signal and high H2K27A signal in liver. So if we click this CRE, we now are brought to the CRE details page. We can start investigating it further. We see overall the CRE itself has very high signal in liver tissue. So liver here, right lobe, as well as intestinal tissue. It also has really high DNAs in colon. We can actually look at intersecting transcription factors as well. And what's interesting is a lot of these factors are almost exclusive to the liver. So for example, a Fox A1 here, if we click, we get brought to a factor book. We can see that in its title, it's actually a hepatocyte nuclear factor, but it's also a transcriptional activator for liver-specific transcripts. Additionally, even in the more ubiquitous transcription factors, we can see that all of the experiments are done in liver. So this is a very liver-centric candidate regulatory element. If we look at associated gene expression, we can see once again we have really high expression in liver. And this is nice because not all the cell types are the same between gene expression data and CRE data just based on what's surveyed in ENCODE. So it's usually another validation. We also have rampage signal, once again, very high in liver. This particular CRE also has an orthologous CRE mouse, which you can click and see. And no surprise here, it's also active in liver. So it's nice that you can use both the human and the mouse data together to further validate your results. What's interesting, I'm just going to go here to this last tab. It's called link genes. So we've been really trying to work on linking distal regulatory elements with their potential target genes. And this is what we've incorporated so far, which is 3D chromatin links through Chia-PET data as well as genetic links through EQTLs, primarily from GTX. So here we actually can see that this CRE has a SNP that is linked with this ABCG8 gene. But we also have supporting Chia-PET data for ABCG8 and it's nearby gene ABCG5. So it looks like since we both have genetic data and physical chromatin contact data, it's most likely regulating this gene. And we can see this further if we go back to our page here. And we want to actually view this in the browser. We can search, for example, liver here. We can pick the right lobe. And we can actually view this in the genome browser, which we'll load in a second. So because UCSC is so bogged down sometimes, this is why Michael's been working really hard to incorporate the genome browser in screen itself. But we can see here, we have on top, we have all of the cell type agnostic CREs. Then we have ones that are active specifically in liver along with the corresponding tissue. So we can see that this distal CRE actually lies in the ABCG8 gene. But in most likely context, these two genes here which have bidirectional promoters. So it appears that biologically that this step would most likely affect the expression of both of these genes.