 So step one is to understand biology. And in the case of a virus such as COVID-19 that's going to be about all the proteins that we have in this molecule. In theory, you could target the things that are not proteins, but I won't really talk that much about it. 99% of everything we do, I would say, is about targeting proteins. In this case, we know what the proteins are because we have the blueprint. How did we get the blueprint? Sequencing. People have sequenced COVID-19, so we know exactly what it consists of. Let's have a look at that. I've taken some slides from another presentation I gave. So COVID-19 or SARS-CoV-2, which the virus itself is called, consists of almost 30,000 base pairs. And it's a single-stranded RNA virus. You might recall me imagining that. This is expressed in a bit of a special way, as you actually have one long chain that expresses all the proteins in this virus. And then some of the proteins in these chains end up being cut off. And that protein itself is then responsible for cutting up all the other proteins. So that's called a protease. That's the blue one here. That's going to be the papain-like protease. And actually you have a 3Cl protease too. Do you see the yellow one up there? That's the spike protein that we're going to cut off, and then that will appear on the surface of the virus. So in this case, there are a handful of different proteins here. In principle, we could target all of them. But if we don't know anything, we would need to start by targeting the ones that we believe are critical for the biological process. So which ones would that be? Well, the protein that is responsible for cutting up all the other proteins, those two proteases, right? If we remove those two proteins, the virus is not going to work anymore. So they are definitely interesting targets. Then we also have that spike protein sitting on the surface and being responsible for recognizing and binding to cells. That also sounds like it's very important. Then there might be some other minor channels or so. And what we usually try to do there is that we look at related sequences. So if something occurs in very conservative form for virtually all of these viruses, it's likely something important that we should target. And vice versa, if there is one of these small proteins that doesn't even occur in some of the viruses, apparently it's possible to live without it. That's not a good target for drug design. So when SARS-CoV-2 appeared, it was only a matter of weeks before this had been sequenced. The sequencing itself just took a few days. And then it was also a relatively job using bioinformatics and phylogenetic trees to start to map this out and realize that this was a SARS-like protein and it was related to other coronaviruses. And that meant that we already had a bunch of information. We even have structures of some of these proteases. The big question then is how similar is SARS-CoV-2 to the previous ones? Well, just looking at the sequences, most of these were way above 30 percent sequence identity. And then you know, based on the bioinformatics lecture, what you can do right, you should be able to build homology models. And people started doing that immediately. I'll show you some of those homology models later because they're done good. On the other hand, given the impact of this virus and that there were thousands of people dying in China, it's no big surprise that there was also a number of groups that immediately started to purify samples and took this into a lab to determine cryo-electron microscopy structures of them. I'll get back to that in a second. But the reason we need those structures is that if you want to target this system, I need to understand exactly how it works. And then we're back in the central dogma, right? Sequence here leads to structure, cryo-in, leads to function, viral infection, or in our case, the function that we want to be able to alter. The proteins that we primarily try to alter is on the one hand those small proteases, but also the spike protein, the spike protein on the surface. As I think I already told you in the bioinformatics, no, the membrane protein lecture, that the spike protein binds this ACE2 receptor, which occurs, say, in our lungs and blood vessels and everything. It's a receptor related to blood pressure, actually, that I'm going to tell you about for another drug, too. So given that we have this interaction between them, the next step was to see if we can determine some sort of complex. Can you determine the structure of first the spike protein and then, ideally, the spike protein bound to this protein, which appeared and some of these appeared very quickly. I think it was a month or so before we first saw the first structure of the spike protein now from SARS-CoV-2. The reason why this was so fast is because there were similar structures available for earlier SARS proteins. And once you had that, people also determined structures when you see that the small domains up here, here we have a version of it bound to where the receptor binding domain is bound to the human protein. And so we're here, we're in business, right? Because now we have at least one of the proteins, we have a structure of it, we have a rough idea of the biological process, how it's binding. And at that point, we can start to go after this process and see, can I knock out this binding? Can I change things so this cannot bind to my cells? How are we going to do that? We will see shortly. There are a bunch of other proteins that we determine the structure of here. Actually, no, sorry, this is another spike protein structure. And in particular, when it's undergoing a structural change in the entire complex with the AC2 binding domain. This is the other one I wanted to tell you about. This is a structure of one of the proteases. So again, this is one responsible for cutting things up. The reason why we were interested in the proteases is that they have been quite successful at creating inhibitors for proteases and other viruses. For instance, HIV1 that I will also talk about later. Given the success for HIV1, it's obvious that we might want to target this for SARS-CoV-2. From the side, this is actually a dimer. So maybe we could just disrupt the dimer formation of this molecule by putting something in the middle here. Again, you will see that for HIV1 later. And that could at least in theory disrupt the protein. So why aren't we just doing vaccine design? Well, as you've seen, people are doing vaccine design too. But these are two kind of parallel efforts. First, at the start, we didn't know whether vaccines would be successful. Now, I would say they are. It's still uncertain how successful vaccines are going to be against mutants, but it's likely that we will be able to develop new vaccines. However, the problem with vaccines is that they're only going to work if you're immunized before you're infected, right? There is nothing you can do with a vaccine if you show up in the emergency room and already have COVID. The idea with these small drugs, so-called antivirals, is that we would be able to treat you even after you have an infection. So you have been effective with the virus, but could we go in and administer these drugs and stop the infection? And in general, that's of course how a whole lot of other diseases are treated, not everything is a virus after all.