 Welcome to the Baxel webinar number 56. Today's presenter is Adam Hospital for IRB Barcelona, Spain. And he will present about 3DRS, a web-based tool to share interactive representation of three-dimensional molecular structure and molecular dynamics reactor. I will host this webinar together with Stefan Fara from the University of Edinburgh. My name is Alessandra Villa and I work on the Royal Institute of Technology. So, Adam is a postdoctoral fellow at the Institute for Research in Biomedicine in Barcelona. He is also research software engineering in the Spanish National Institute of Bioinformatics and is leader of a workflow package in BioXL Centre of Excellence. He has developed a series of public web server and database, for example, an MD web, a nucleic acid flex and a big nucleic acid seam. And now he will speak about the new tools that he has developed. Please, Adam. Okay, fantastic. So thanks, Alessandra, for this nice introduction. As Alessandra was saying, my name is Adam Hospital and I'm going to present today a tool which has been developed in the Institute for Research in Biomedicine in Barcelona. Within the BioXL Centre of Excellence and the tool is called 3DRS, 3D Representation Sharing, and basically is a web-based tool to share interactive representations of 3D by molecular structures and also molecular dynamics trajectories. I've divided the webinar in four main sections. I will give you a brief introduction about the tool, about the idea behind the tool. I'm going to try to convince you of why we have developed this tool and I will show you similar tools for you that I think are really useful tools in our field. I will explain a little bit of how it works, including the main functionalities, how you upload structures, how you edit the representation and how you share the final representation with your colleagues. I will give you a live demo of the web server and I will end up with some conclusions. So starting with the introduction for the tool. This is a true story. It all started when I was reading a paper, not this particular paper. This one is pretty recent from two months ago. But I think that it's a really good example that illustrates very well the idea behind the tool. And this, for example, this paper is demonstrating how a glycan gate controls the opening of the SARS-CoV-2 spike protein. I'm sure that all of you now already know what the spike protein is from the SARS-CoV-2, but basically is the protein that makes the SARS-CoV-2 virus able to recognize the human host cell. And it does so using this small domain here, which is called receptor binding domain from the spike protein. And in order to do that, it needs to make a kind of a conformational transition, needs to detach and get exposed outside of the spike protein. That means that in this case you can see here a conformation that is completely closed and slowly it is opening till the last conformation which is completely open. When it is exposed, it is able to recognize the human cell and it starts the infection. You can see here also that there's a couple of views, side view, top view, side view and top view from the closed conformation and also the open conformation. I'm showing you that because when I was reading that or a similar paper, it always came to my mind that I really would like to see this kind of static figures in a more dynamic way. I know that there's a lot of different conformations here as you can see, that we can see maybe in a kind of a movie. Also here, top view, side view, we know that we have interactive tools to do that in a more interactive way. And I'm always thinking about this kind of future journals, if you want, from like these ones that appeared in the blockbuster movies in the beginning of the new century, you know, Harry Potter, The Daily Prophet and Minority Report USA today. Of course, we are not there yet in terms of technology, but what we can do maybe is to work a little bit with HTML pages like this one. And maybe we are not talking about the future kind of journals, but we are talking about more recent journals and I will try to convince you about that. Actually, when I was looking at the, I was reading at the paper, I was not reading physical paper, I was reading a PDF like this. I'm sorry, a PDF, an HTML file I was reading directly in the website. And I was scrolling up and down and looking at the figures, as you can see in this GIF. And basically this is something that we are kind of used to that now. Actually, the new generations, they are grow up, knowing how to, how to work with a touch screen, how to click, how to scroll up and down. And this is something that is not the future anymore. This is the present. So we started to analyze, if we want to convert this kind of static figures to something more like live figures or interactive figures, what should I, what should we do and what should we, what do we need for that. And the first thing that we need, of course, it's a three dimensional representation or viewers like these ones. I'm sure that you all know them. They are really famous nowadays, but they were kind of a jump, real jump between the old technology and the new one because they are all working with WebGL libraries. And these are the ones that they are using GPU power, graphical processing unique power from our computers, jgsmall, ngl, molestar, lightmall and this many of them that I'm sure that you know. So this kind of HTML embedded viewers are present. We can use them. And actually we are, we have been using this now for for one of them then GL for years in our website so technology for this part is ready. Interactivity in websites in HTML pages. I think that it is ready as well. You can see here one of one tool that we developed a couple of years ago in the group that basically it's showing you an analysis and one analysis of a long DNA fiber and you can see how the three different sections in the website are completely linked one with the other. This is an ngl viewer that you can zoom or rotate wherever and and you can click on one of these, for example, plots and the particular structure is represented here so we have the interactivity. We have also the 3D viewers and what we also have and this is because we are now in the era of the streaming data. We have a really high bandwidth nowadays. You know, we are all watching series and films and movies in Netflix and kind of streaming platforms and the pandemic also helped a lot on that. And basically, we have everything that we need to try to go from static figures to a live figures we have WebGL tools we have interactivity, and we have a network one. So, can we add, can we try to add interactivity and movement to this kind of static figures and go from the, you know, the future journals that I was introducing before to the present journals let's let's try to do that. Now, using HTML websites. Spoiler alert, we can. And I'm going to show you. Let's see if that works and I am now changing from the presentation mode to an HTML browser. And I'm showing you our publication our paper for the 3D RS tool and if you scroll down here this is HTML again. You will see static figures like this one this figure one and you scroll down more static figures. Another one, but when you reach the results page. Now, this is not a static figure anymore this is a dynamic figure that you can rotate you can zoom. You can see for example in the spike protein the different domains and terminal domains are bd for the three three different chains. You scroll down a little bit. You can see here how an antibody is attached to the receptor binding domain of course you if you want to see that in a little bit more detail you can make it full screen. You can rotate it again. You can see for example the red dots here which are the residues from the South African variant of the virus have many information in the caption of the figure as if you were looking at the figure but in this case interactive figure, not static figures anymore. And if you scroll down a little bit more. You can even see something like this. There is not just a figure that you can interact with but this is also a movie. And in this case, again, a movie of a project related to SARS-CoV-2. This is what you are seeing here is the RNA dependent RNA polymerase from SARS-CoV-2. This is the mechanism of reaction and inhibition of Remdesivir and here you can see how a bond here is slowly broken and another one is forming. You don't have time to maybe see that but I invite you to take a look at the gallery of 3D RS to see this nice movie. Going back to the presentation so we can do that and at that moment we sit for a while and ask ourselves if this was worth developing it just to have live figures. I can tell you that in my opinion this was worth it because I really like the idea. But we also thought about why if we can have something like this which is not just a live figure but also a URL to share the representation with our colleagues. And this makes a lot of sense, for example, for internal meetings. It happened to me a lot of times that I went to present results to my boss and presenting him. For example, taking this figure as an example, we have a two magnesium, the coordination of two magnesiums here. And in one particular snapshot of the MD, something happened and my boss was asking me, okay, I want to see the structure. What happened with the structure? What happened with the coordination? What happened with the atoms and the residues behind it? And I had to look for the trajectory, open BMD, try to build the representation. It took me like 5 to 10 minutes and my boss was no happy with that. And this is even worse when you think about collaborative projects, you're not in the same room. And as an example, think about our bioxial center of excellence with the different European partners when we are sharing information from a particular MD simulation or a structure or model of the structure. It's really useful to have something like a representation already built and a URL to share the representation. And finally, also for representations like this one for webinar for conferences, one thing is to show one particular representation like this one or a video. And another thing is, if you can show the representation and also share the representation with the audience, with just the URL that you can share. And this is, for me, it's really interesting. So, of course, at that moment we thought, okay, let's try to look at the literature and see what is already there. Tools that are already existing with these functionalities and we, of course, we discovered that there was many of these tools. Proteopedia was maybe one of the first ones to integrate the animations of structures and now they are using JS mall. There's also PolyView 3D integrated in a framework of bioinformatics websites. You have this nice IC in 3D viewer that is integrated in the NCBI websites, also 3D bio nodes from Elixir Spain. None of them were really focused on generating representation in a really easy way and then sharing the representation. Although in most of the cases you can have a URL at the end to share the representation, but they were focused on linking the structure with different sequences databases with analysis with any kind of tools. They are really complex, many menus with 1000 different options. That was not exactly what we were looking for. We're looking for something like Web 3D mall. It was the most similar to what we were thinking at that moment. But you see that this is easier, at least in a graphical interface than the other ones. But it was missing, sadly, a couple of functionalities that for us were crucial. One of them is we wanted to be able to upload the structures models, the structures, theoretical structures, snapshots coming from molecular dynamic simulations. And in this case you can only use PDB structures coming directly from the protein database. And of course the other functionality is the possibility to add molecular dynamics trajectories that was crucial for us. It was not implemented at that moment in any of these tools. So we thought that maybe we can start developing our own tool and we started with the 3D RS development. How our 3D RS works, these are the main functionalities of our tool, five main functionalities. It is important to highlight that one of the main objectives was to make an easy tool to generate 3D and also 4D by molecular representations. And for that we used state-of-the-art technology, single-page application, web-based graphical user interface. That means that with a single web application what you have is a seamless user experience. It's like that you are working with a standalone program, no reloads, no saving buttons, you will see that in the demo. And to make things even easier we used the BMD-like selections and representations. Anyone familiar with BMD will understand in just two minutes how to generate new selections and new representations with this tool. And you will see again that the tool is able to represent and show molecular dynamics trajectories. And for that is using MDSERF technology. It's able to not just represent one single structure at a time, but you can upload multiple structures at the same time. And again another of the main objectives of the tool, it generates a persistent URL to share the by molecular representation with your colleagues. After that we added a new functionality, which is the representation forking. And you can think about that as the forking that you can do in GitHub repositories for software. And that means that you can generate a copy of the whole representation and then start modifying your own copy of the representation. And this is how it works. It's a cycle of 3D-RS. You start editing your own representation when you're happy with the representation that you have built. You can share the representation and the final user. If she wants to modify the representation, the representation can be forked and then they have their own copy of the representation that can be extended, modified in the addition and then shared again. I was highlighting before that the graphical user interface is really easy. And here this is an example. There is a launch section, launch menu here. And you can launch from a PDF file or you can upload your own structure or structures, drag and drop really easy. And after uploading your own trajectory, you can start editing. This is the cycle again. This is the addition mode of the tool. And this is the sharing mode of the tool. You start with the addition when you are happy with the representation, you share it and you enter into the shared mode with a persistent link. If you want, if the final colleague that is looking at the representation wants to fork the representation, it makes a copy. With the fork and starts again with the addition mode, but starting from the representation that has been shared before. So I will explain a little bit about the addition mode now. As I was telling before, it's the selection and representations, they are VMD-like. That means that just a reminder for all the ones that are familiar with the VMD and also for the ones that have never used VMD. You will see that this is really easy. You have the idea of representation. You can create new representations here. All these lines are new representation. If you double-click the representation, the representation is hidden. If you double-click again, the representation is shown. You can of course delete. Every time you add a new representation, you can modify the coloring method, the drawing method. Of course you can modify the selected atoms for the representation. And this is mostly everything in the selection and representation part. And then you have here the possibility to add a molecular dynamics trajectory. Here you have a number of frames. Here you have a control player that you can go forward and backwards. And you can modify the step and also the speed. So this is exactly what more or less what we reproduced in our graphical user interface. Here you can see our addition mode and it is divided in different blocks. The first block is the representation block where we can create a new representation and modify the molecular representation and the color scheme. And in the second block, we can select the fragment of the structure that we want to represent with this kind of representation. So both windows are completely linked. Here you select the fragment of the molecule and here you represent, you change the representation. Of course in the middle you have the stage panel. You have some useful tools here in the toolbar. And you have the final button to share the representation when you're happy with the representation. In a little bit more detail, you can select here in the selection window. You see here the selection window is the right part of the graphical user interface. You can select by chains. You can select by residues here from the sequence of the, in this case, the protein. You can select also heteroatoms, ions, waters, all that is included in the structure that you have uploaded. You can also have the possibility to click here and make a custom selection, and this is using NGL syntax. If you click here, I will do that in the demo. You will see that NGL has a really powerful syntax for selection of particular fragments of the structure. It is quite similar to BMD again, so it's easy. You have also the possibility to select clicking on the particular residues or different residues in the stage. So if you click on a particular residue in the structure represented in the stage panel, it is also selected and added to your selection that you are working with. After having all the selection ready, then you can modify the molecular representation and the color scheme. The molecular representation and color schemes that are available are the ones that you can imagine, the ones that are also integrated in BMD, and also in NGL, backbone, ballistic, cartoon, you know, all of them, and also the color scheme, by chain, by B factor, by element, the typical ones. You have also these small buttons here to hide the representation, such as in BMD, edit the representation name, clone a particular representation, and open the label settings. You have here the label settings that you can modify the name and the size and the color. But this is everything. There is nothing else than that. It's really easy, believe me, and I will try to convince you during the demo session about that. The last thing, the last block from the demo mode is the toolbar where you can, well, you have useful functionalities such as reload the scene, center, change the background color, go full screen, superpose the different structures, modify the camera type from perspective to orthographic. And as I was telling you at the beginning, it has also the possibility to upload trajectory files, in this case, MD files in these formats, and it is done thanks to the MD serve MD trajectory server that is implemented behind the tool. The demo 500 megabytes size, of course, you need to reduce a bit your trajectory if you want to share it, but we have tried with many different trajectories and I can tell you that 500 megas. It's enough to represent a trajectory nowadays, of course, reduced. Of course, you have trajectory settings, you can change the range of the trajectory that you want to expose. You can change the step the interpolation from one frame to another, the timeout which is basically the speed of the trajectory. You can make the trajectory loop or rock, and you can make the representation when you share the representation, you can make it out of place so the dynamic simulation will start playing directly without the player button. And finally, there is the sharing button. And this is important once you're ready, and you're happy with your representation you click on sharing and you can take a look at the final draft and if you're happy you can click on share the project. You can enable fork, this is important. If you enable the fork, your colleagues will be able to continue copy your representation and modify and add or extend the representation always making a copy. If you deactivate this, then there's no possibility to copy the representation, you can make the project public or private, if you make the project public it will be included in our list of projects in the homepage of 3DRS so that all the community can take a look at the different representations. When you click on the share project, you will have this short URL which is basically the link that you need to share the representation. You have also the possibility to copy this embed HTML snippet in your HTML and this is what allows you to make these live figures, such as the one that I have presented you at the beginning of the presentation with our live paper. And you have also the QR code if you want to try it's nice to use the smartphone to look at the representation of your macromolecule. Remember this is a cycle. I have explained you the addition and the share. We are now going to take a look very briefly to the shared mode once you have a persistent link what you can see. The main difference between the addition and the share mode basically is that in the share mode you cannot modify the representation. So the representation is there as the author of the representation wants, including the zoom including the rotation of course you can then modify the zoom and the rotation but you cannot modify the selections, the representations, the labels, the background color, anything like that. So it's a shared mode, it's not an addition. You can still interact with the scene and you can still see the atom and residue information, this small box that appears at the bottom of the interface when you hover over one of the atoms of the structure. What is new in this case is the figure caption, the trajectory player if you are sharing an MD representation, specific tools for the toolbar and the forking process. And this is what I'm going to tell you now in just a couple of minutes. Those are the blocks for the shared mode, you have the stage panel again, the main representation you have a figure caption, which reproduces a caption of a figure in a paper, but in this case a little bit more interactive so you have bold and italic type of letters you have also links that you can also of course add to the figure caption. You have the trajectory player here with a slider here that you can control, you have a play forward and reverse, you have the toolbar that in this case is a little bit different, you can take a screenshot, you can get the embed code to add it in an HTML, but you cannot modify the background color for example you cannot modify the perspective. And finally you have the forking button here. This is almost everything that I wanted to tell you about the 3D areas now I think it's time for a demo, which is always nice to see how it works in a live session. I have here prepare one PDV ID remember this is the launch menu from the tool. I'm going to put here one PDV which is the EGFR file, epidermal growth factor receptor, the kinase domain. In this case this one as you can see here is the one that is complexed with Kefitini before not the FDA approved drugs, you know the EGFR is really popular, because it's used, it is related to many cancer diseases and this is one that is used in these cases, Kefitini. You have a default representation when you start using the tool. It's this default representation here it cannot be changed it's just for you as a to have a reference of everything that your structure contains here you can see that there's water molecules there's a sodium ion. There's the Kefitini small molecule there's the protein everything you can hide the representation anytime you want, but I recommend you to use this representation as a reference you just can modify the opacity the transparency and keep it there. And then you can start creating new representation this case I will create the new Kefitini representation for the small molecule. When you started representing there's a lot of information here saying that there's a new representation created. You can close that you have the new representation which is now now contains everything all the different. Atoms of the structure but if you go to the sequence, you can go and select any residue that you want for the protein you can select any other atom that you want you see that is highlighted in the stage panel. You can also select the ions in this case the sodium ions and finally you can select the waters. You can select one by one or you can select all of them in this button here, you can view the sequence in an extra window here is the protein you can also select by secondary structures. But in this case we are interested in selecting the at the bottom this one here I can center the view this button and I can use the click option to select now. It says that these at the bottom structure has been added to the representation so now we have the representation we can take a look at that. We can modify sorry we can modify the molecular representation to licorice now that it's just this small molecule that is represented give it a little bit more of radius so that we want to highlight it a little bit more and modify the color scheme element. You can always go to the default representation and deactivate and see exactly the representation that we are working with. Okay, I think that now I have the representation that I want. I'm going to now generate a new representation I'm here a GFR representation. And in this case what I'm going to do is to select all the residues of the protein just the protein I can click here and select all the residues all the residues has been selected. And now I'm going to modify this to cartoon. And again, if you go to the default and deactivate that you see that there's no waters there's no audience just the protein and the legal. It's okay then I go back to the protein I will lower down the opacity because I'm really interested in the active center and not just a protein. So now I'm going to create a new representation active center. And now I'm going to use the custom selection. And I'm going to say okay I know that yes if we go here. You can see the syntax language selection language of angel so you can see that is really powerful there's many different options that you have one of the options to select is using the residue name. So, I will use the resume name to select the get fit in it that I know that is called it. I are you see in the bottom of the page you always have information from the items that you are hovering. So I select that. And actually I'm not interested in just this one I'm interested in the active center so all the residues that are in a certain distance from the legal. So five Armstrong's in this case I can modify that but it's okay. I am interested in the whole reduce surrounding this. And now I have the representation of the surrounding residues I want to modify the representation to ball and sticks to element I kid who. And I will now reduce a little bit the radius because I'm interested in seeing all the different atoms. Okay, I did something wrong I selected something that I shouldn't let me do that again. Don't worry. I just wanted to if this happens there is a button here which is really nice which is enable navigation mode because that makes you let's you navigate through the scene and click on the different. Atoms and residues without affecting the selection. And, for example, you have water molecules. I hope that you can see the water molecules here that are selected as being on the active side. Okay, now I'm going to. Let's disable that to clone. What is the clone clone the current representation because I want to have the same active center but with another representation in this case I want to represent it with surface. I'm going to put here a uniform color and something like, like this. And again I will play a bit with the opacity and transparency I think that's enough and this way I have the active center highlighted I have all the residues from the active center. I have a small molecule and I can see the different interactions. And this is nice because you actually can interact and can rotate and consume when you share the representation the only thing that you have that I have done is to highlight the active center. Okay, that's enough. Now I'm going to add a caption with the project settings. This is the bio Excel webinar. EGFR protein, Excel, 3D RS webinar here maybe. EGFR protein, the pdb code is the 4WKQ. And I can add a link that I have here, just like this. And I will make this bold for example. This is saving and you are not clicking any button to save. And you can close that and now we are going to the share button and let's see the draft, the final draft. So this is the final draft that you're going to share if you want. You have here the caption that you can expand. Of course you can rotate. You can zoom. You can go back and reload the representation. You can spin. You can make it full screen. You can take a picture of the representation. And if you're not happy with that as a final user and you want to, I don't know, take a look at the different interactions here or make a bond. I don't know if we have time for that. No, don't have time, but you can fork. The project, well, this is a draft, sorry. I need to, this is a draft. You need to enable the fork and make the project private in this case. I can share the project and now here you have the final short URL that you can copy and share. If I open this one, this is a final representation shared. And now in this case I can fork the representation. And if you fork the representation, you basically go back to the addition mode. And here you have everything. All the different representations that I have created. You can modify, you can add more. And remember this is a copy of the original representation and you can make your own representation. Okay. I think that's all for the demo. Let's use the last five minutes to go through the conclusions. I just presented you a tool, 3DRS, that it's making easy and this is important, the generation and sharing of living interactive 3D and also 4D with trajectories, custom representations. It is designed with the state of the art technology, single page application. It has these selection and representations, VMD-like, so they are really easy to understand and it works with NGL behind to efficiently show the 3D representations and using MD-Serve to efficiently stream the monochromatic data. I invite you to go to the gallery and take a look at the different examples that we produced. They are really, I think they are really nice. You can fork them, all of them are forkable and you can take ideas from that and from these ideas. I want you to try and play with the tool. And as I will tell you in the next slide, if you make this public, this will be even better. Remember that you have extended, really extended documentation in read the docs that you can take a look with all the functionalities, shortcuts, keyboard shortcuts. It is much more than what I can explain in just one webinar behind 3DRS. And what I was telling you one minute ago, please, when you try and play with this, if you can make your project public, this will be added to our list of projects in the homepage of 3DRS so that all our community can take a look at your representations and see what you can do with the tool. And of course, any type of feedback would be really welcome. I mean, European comments, results, suggestions, ideas, all of that will be really welcome for us. With that, I would like to close thanking, of course, to my group leader in the molecular modeling and bioinformatics, Modesto Roscoe from Ayeri Barcelona and to the very, very talented full-stack developer, Janice Bayari, who did everything that you have seen in this webinar from the database that is behind that to the installation on the MDSERF and all the technology and also the graphical user interface. And I would be really happy to take questions for you and discuss about 3DRS. Thanks for being here and for your attention. Thanks for a great talk, Adam. It was very nice, very clear. Yeah, so we'll do the Q&A now. So some of you have written in questions, so I will read them out. So if you type a question, I will read it. If you want to speak a question, then I think we've now enabled the raise hand feature on Zoom. So if you click raise hand, then I can unmute you and you can say your question out loud. So it's up to you. So the first question we have is, so where is the information stored when we share a persistent link? Do we need to upload the trajectory via the internet to your servers? Yes, well, there's two answers, I think, to this question. The first one is all the information is stored directly to a Mongo database that is behind the tool. That means that we are basically everything is directly saved in the Mongo database while you are interacting with the scene. As you have seen, there's no save button, so everything is automatically saved there. This is one thing, so everything that you need to reproduce the representation is there. With the exception of the structures and the trajectories, structures are saved in disk, in usual disk, because we don't want the structures to be replicated when you fork the different scenes, so we are using links to the same structure and trajectories are saved to this MDSERF application, which again, it uses a disk, but it is a usual disk, but it is streaming the MD trajectory frame by frame. That's how it is so efficient to use it in tools such as this. This is MDSERF that is doing that. I hope that I have answered everything. So next, do the forks work like GitHub? So can you do a pull request if you change something? Not really because actually you're not changing the master in this case, you're not changing the original, but you are making a copy. If you are forking in this case, it means that you're making a copy of the representation, and after doing this fork, you are working with your own representation, starting from the representation that it was generated before. So no needs of pull requests in this case. So can a user extracts data, i.e. the structures or the trajectories from the 3DRS representation? No, and there's a reason for that. And basically it's because we didn't want to do that. What we wanted to do is to share the representation and not the data. We are working in different tools to share molecular dynamics, simulation data, where you can extract a particular frame from the trajectory. You can download the whole trajectory. You can download using the rest API fragments of the trajectory, but this is not the tool for that. This tool is just to share representation. Okay, so next one is about, so I think not having private projects is a great open science stance. Could you comment a little bit on the data storage policy? So where is it kept and for how long? Yeah, we thought a lot about this. All data is stored now in our own premises. This is a Mongo database with horizontal scaling. You know what that means that we can add new machines and we have replication, we have sharding. That means that data is kind of secure there. If we lose one node, one disk, we can still recover everything. And after that, we also have backup policies. But this is in our own premises. And the idea is, if this server is popular enough, if we have many, you know, if it's successful, let's say, then we can think about moving all of that infrastructure to a more robust infrastructure. Let's say BSC, for example, because we are working with them. But now it's everything in our own premises. It's, you know, it's secure for all the things that I have told you, but we are thinking about moving if needed to a more robust infrastructure. Yeah. And in the time now we don't have time. Well, you've seen when you create the representation, if you don't share the representation, it is removed from our server. We are not saving everything. But once you share and you generate the persistent link, we don't have expiration date now. So that means that we will keep that here happily forever. So can you render when you save the figures? So I guess like when you render with VMD? Not for the animation, but yes for the representation. So you have a take a screenshot that comes with the NGL. There is nothing that we have programmed. The NGL has this functionality. What we can modify, well, what we modified is the resolution. We have put a very good resolution so that you can take a very good screenshot from the representation. But there's no, as far as I know, there's no way to extract movies from the animations. If one protein does not have a PDB file, then can I use Alpha Fold 2's prediction of that protein's PDB format? After that, can I use it in 3DRS? So I guess. Yes, as Alpha Fold 2, if I'm not wrong, it is giving you PDB format files. You can use this PDB to upload this in 3DRS. I was thinking we have not integrated the PDBs generated from Alpha Fold 2, but we can do that. We have done this in another tool. Maybe it's something that we can think about. So can the representation be embedded into another web pages? Yes, you can do that with these embedded snippets that I was telling you during the demo and I think also in the presentation. Well, if you want more information about that, there's in the documentation, but when you share, when you click on the share button, you get three different ways to share the representation. One is a URL, short URL. Another one is HTML to embed, a snippet of HTML to embed that in your HTML website. And the third one is the QR code. Okay. So I guess, so this question says, is there a cap on the file size? I think you said was 500 megabytes for the trajectories. Correct. So is that something you're going to sort of keep or I guess, does people run longer simulations and bigger ones? You might need to. Yes, but yeah, I also commented that, of course, we have much, much larger trajectory files than 500 megabytes. But if you can reduce this structure to something like my 500 megabytes, you can reproduce very nicely, I would say trajectories that are now generated nowadays. So are any of these interactive figures allowed to be published in papers at the moment or I guess are you in, have you been talking to any journal websites who would allow you to embed these HTML links? That's a really good question. But well, the first idea was exactly that to talk about that to the different journal editors. We have started the process not really successful so far. It seems that it's complicated to jump from now to the future. You know, I think that we will go slowly go there. You can take a look and you will find in the literature different journals that are accepting some kind of interactivity live journals. They are slowly appearing, but in our case, it is 3D RS is not yet implemented in any of the journals, but we will fight a little bit for that. I guess this is a similar question. Is there a feature to represent these visualizations in presentation software? So I guess that sort of PowerPoints or Google slides. No, there's no way. I had to make gifts of all of that. Again, I think with HTML you can do almost everything nowadays. I'm sure that the embed code will be, I don't know now maybe there's something and we will need our main developer for that. He knows, but if it is not there, I'm sure that in months from now we will have something to use this embed HTML and use it as a presentation. I don't know, but I'm sure that there will be something soon. So can net CDF trajectories be represented by this tool? I guess to an extension to that, what file formats can you supply molecular dynamics trajectories in? To be honest, I don't remember, but if you let me go back to the slide, I guess that you're still seeing the presentation. Yeah, yeah we are. I have one for MDSERF somewhere, sharing and trajectories. And you can see here it's really small, but is XTC, DCD, TRR, BIMPOS and net CF files are accepted. Again, this is not something that we prepare. This is something that is in the MDSERF technology. But I think, you know, nowadays you can compare from one to another very easily, so it shouldn't be a problem. So then we have another question. So what's the role of the server in your implementation? The rendering appears to run client-side, couldn't the whole thing be serverless? Well, I guess it could be. The only thing is that we have to use, you know, the new functionalities from the new web browsers that gives you a kind of database, client-side database. But I think that, well, we discussed about this and we thought that it's more secure and it's clear to have something in server-side. We have control of our Mongo database. It can be really scalable. And of course it's more flexible than using just the memory of your browser that is limited, of course. But yeah, that's a nice suggestion. We can explore that. Nice. So that's what's going on. We have some more questions. Which is good. So is there a plans to implement a gif maker feature, a gif maker feature so you could easily generate gif images from representations using 3D RS? Not now. Well, I was thinking that if you have an extension of Chrome browser or something like that to generate the gif is as easy as to use that. I don't know if this will add value to our server, but we can explore this depending on the difficulty of that. If it's really difficult, I will be happy to just forward our users to the gif maker as an extension of any browser. Okay. So I have a question actually. Have you sort of, so I guess there are lots of web browsers and lots of machines out there. Have you tested which web browsers generally work? We have tested with the more popular ones. You know, Firefox, Chrome, Safari, we had some problems with Safari, but we solved them. We still, we still know that there's some issues with particular browsers. We have also tried that. So I think that we are covered with the most popular ones. Okay, so I think that's all the questions we have now. So if anyone has anything else then feel free to send them in or raise a hand. In the meantime, maybe we can announce the new webinar. Okay, so the next webinar will be the 9th of November. And that we will have Charlotte Dean, and she will speak about computational design, therapeutic, anti-building combined in moon report, data and structural information. So you are welcome to register in the BioXL from the BioXL page you find the link to the webinar. And if no other questions are popping up, I would like to thanks again Adam for the wonderful presentation and all the attendees for attending and for making a lot of active discussion. Thank you very much. And we close here the webinar. Thank you.