 Hello, good afternoon and welcome to the next webinar in our webinar series. Today, Federica Battestini is going to be talking about NAFLEX, a web server for the study of nucleic acid flexibility. My name's Adam Carter, so I'll be hosting this webinar, so I'll just use this time for a couple of minutes to tell you very briefly about BioXL for anyone who's not familiar with the centre and what we do. And then I'll hand over to our speaker today, who will speak for the rest of today's webinar. Just a quick note to let you know that this webinar is being recorded, including the Q&A session at the end, so you should be aware of that. And we will post the recordings afterwards on YouTube and the BioXL website. So for anyone who's not familiar, BioXL is a new centre of excellence, and we've got three sort of main areas in which we're working, sometimes referred to them as the three pillars of BioXL. One thing this centre of excellence is hoping to do is to improve the performance and efficiency and scalability of some key codes that are used to do biomolecular research in Europe. In particular, we have developers from the Gromax code for MD simulations, from HADIC for docking, and also from CPMD people working on the QMM interface for this code. So the key codes are part of what we do, but another important aspect is usability. So as well as improving the software in terms of functionality and performance, we want to make sure that all the codes are usable. And an important aspect of that is how they can be used as part of wider analysis workflows, and workflows are something that is key to the project too. Finally, we have a work package or a couple of work packages involving consultancy and training, so we very much are interested in hearing from our end users and what you're working on, what you're interested in. That helps to define the direction that the centre of excellence is going in. To that end, we have a number of interest groups that may be of interest to you. The list is on screen at the moment. So if you're interested in any of these, please do go to our website and you can join up those interest groups and find out more about what we're doing in these areas. We will have our main talk today, which we'll probably run straight through from beginning to end, and you'll have some opportunity to ask questions at the end of the webinar. You can do that using the questions tool in the GoToWebinar control panel, although it looks similar to what you can see on the screen at the moment. Not identical, but you can ask your question here and then I can, either if you have a microphone, open it up and invite you to ask your question to Federica directly, or I can read out your question so you get a chance to, so that Federica can answer your question. Finally, if you're watching this presentation online, so you haven't got a chance to ask your question live, do please go to our discussion forum at askpirexcel.eu, where you can leave a question or leave a comment and we'll get back to you about that. So now I'm going to shortly hand over to Federica. Federica Battistini is a postdoc fellow in Modesto Orozco's group in the structural and computational biology program of the Institute for Research in Biomedicine, that's IRB in Barcelona. She has the background of a physical computational chemist, but she has also worked in the field of biology. Her undergraduate master's degrees were in physical organic chemistry from the University of Milan. Then she did her PhD also in physical organic chemistry at the University of Sheffield with Professor Christopher Hunter. So that was a theory-based approach to understand sequence effects on nucleosome positioning and chromatin. Since 2010, her postdoc position, her research is focused on the study of nucleic acid dynamics at the molecular level and the sequence effects on physical and chemical properties of these biomolecules, as well as epigenetic effects. And Federica has been involved in several collaborative projects with Barcelona Supercomputer Center, and she's also been involved in some of the work that we've been doing here in biorexcel. So she's very well placed to give today's webinar, and I'm now going to hand over to Federica, thank you very much. So if I just make you the presenter Federica, you should be able to take it from here. Hi, Hall. Yes, how are you doing? Thank you. During this webinar, I will present you NFLEX, a web server, sorry, a web server for studying nucleic acid flexibility, isolated or bound to other molecules. Nucleic acids have been, at the first level, are described by their sequence, and in the last decade, many sequencing techniques allow us to have more and more information at the genomic scale. But as polymeric and polymorphic molecules, this info about the sequence is not enough to understand the role in biological and molecular function. For this, one of the information that we need to understand their function is the structure of the polymorphic and polymers of DNA that strictly depends on the sequence. It's not an ideal BNA, but depends on the sequence, and each base pair is characterized by a determined flexibility and geometry. The study of the structure of the DNA will bring us to have more knowledge and allow us to understand many biological processes. For example, the function for protein recognition, protein DNA binding, and the genome positioning of protein and of nucleosome. As you can see here, at the bottom here, you can see the structure, for example, of a nucleosome with the DNA wrapped around the instance. And in this case, oh, sorry, and in this case, you will see how different sequences of DNA will position a nucleosome and either one. And this is related just to the properties of DNA. For this reason, it would be impossible to have experimental data on every possible DNA sequence and in every possible complex. So in absence of experimental approach, simulation techniques are widely used and accepted tools to describe the nucleic acid structure and flexibility. For this reason, we developed and implemented an Aflex server tool to study the properties of nucleic acid in its conformational and flexibility properties in silico. And also to facilitate the use of a nucleic acid simulation tools for newcomers to the field. Here you can see in this slide how is the Naflex is divided mainly in three parts. That is the input part, the simulation engines, and the analysis tools. Naflex is part of a multi-scale genomic project and BioXL. I will go through each phase, but I will focus mainly on the flexibility analysis part. And I will show some results where we directly apply Naflex to solve or to shed some light on biological problems. So starting from the input, in Naflex the possible input is the sequence of the DNA or the DNA in complex with molecules, ligands or proteins from the structure that can be from the protein data bank. So a nucleic and NMR or X-ray crystal structure or from trajectory that have been run independently by the user or from save projects from Naflex. Here I'm showing you what you will see in the web page, so in the web server of Naflex. So the simulation can be run starting, as I say, from a structure that will be the setup and run on the simulation. In Naflex are implemented two ways of running an MD simulation and I will go through the different simulation engine briefly. Then the analysis can start directly from an MD trajectory that can be converted and can be used to analyze all the flexibility and nucleic acid structure in the web server. And can just start the analysis of the nucleic acid can start just from DNA or RNA sequence, then there is a way software implemented to be the structure and from that to run the simulation. Or, as I say, starting directly from a saved Naflex project. What are the simulation engine that have been implemented into the Naflex server? So there are two possibilities. There is the possibility of running and setup of an MD simulation at atomistic level, so with an atomistic representation. And in this case the description of the molecule is very accurate. The problem is that it's very computationally expensive. To allow the user to run an MD simulation has been integrated and directly connected MD web technology that help the user during the setup, the equilibration and the preparation of the input file for the simulation. Here you can see some of the steps for preparing and setting up the minimization of the system, the equilibration of the system, and usually there is the molecule into a solvent box with the addition of ions. If you have any more information about MD web, there is a bioxcell presentation webinar done by Adam Hospital about MD web, how to set up the system and how to prepare an MD simulation. So Naflex is directly connected with MD web and a small trajectory can be obtained or the user can download all the input files and then run the simulation. The simulation, the trajectory can be visualized and then analyzed with all the nucleic acid specific tools. Another option and here I'm going to describe the other option is to define, to represent the DNA with coarse grain models. So you can see here on the left the double strand DNA with the definition at atom level with the backbone and the base pair. So the backbone defined by the phosphate group and the sugar and the bases that form the base pair. Here is the definition of the coarse grain model where the DNA is represented by few spherical bits connected by spring. In this case, in this presentation, the bits are represented by the phosphate, the sugar and the base. So we have two algorithms to analyze a coarse grain simulation that is at the level of the nucleotide base level. That is the mesoscopic acid model and another method that is at M base level. That is the worm like chain model. I will not go into details, but just to show you the two possibilities. One is a Metropolis Monte Carlo algorithm associated to the helical parameters. I will go through the description of the helical parameter, but mainly is the movement between the base pair. And this obtained using an harmonic approximation of the movement of the base pair along the DNA. The worm like chain model used a Monte Carlo algorithm and considered one bit for each four base pair steps. It's important to say that the user can also change their resolution. And here there are some technical details like the Bayesian equation taking into account implicit solvent, ionic concentration. There is a uniform charge of the DNA, but mainly is to consider. We use like Monte Carlo, this Corsair model mainly to describe system that have long chain of DNA where the implication of molecular dynamic simulation at atomic level will be too computationally expensive and will be in with the dynamics that will take too long in time. The first part and is the one that I'm going to spend more time during this webinar is the analysis tool, the part of the analysis tool. That, as I say before, are strictly sequence dependent physical properties. So the analysis of the sequence dependent physical properties. The nflex analysis tool are divided into the standard quantization analysis that are analysis that are standard when analyzing molecular trajectory that is usually the RMSD. So the Routeman square standard square deviation from the starting conformation, a radius of generation, B factor. But mainly in nflex what you will find is a set of different flexibility analysis for just nucleic acid. It's important to say that all these tools, all these analysis that I'm going to go through have a common interface. They don't need additional expertise. There is an easy visualization and there is always an help window for information on the calculation and on the software that have been used. Every analysis can be plotted in along the time, so time course along the trajectory or can be plot just the average value of the analysis that you are running. And have been also integrated in the final results, literature and experimental data that allow the user to have a direct comparison and to check many times the validity of the results. All the raw data that have been calculated can be downloaded. Now I will go through the different tools and I will start with the microscopic descriptor of the DNA and then I will go to the micro descriptor of analysis of the nucleic acid in general or mainly the DNA. Here you can see that the microscopic analysis involves the base pair parameter and now I will explain what they define. And there is defined as Curve analysis because Curve is the program that we are using to analyze the geometry and the stiffness of the DNA. We have analysis of the principal components of the main vector, the motion of the DNA. And then I'm going through microscopic descriptor that can be connected with NMR study, hydrogen bond and distance analysis and energetic analysis at atomic level. Let's start from the microscopic descriptor that are the helical parameters at the stiffness context. DNA can be described at the base level by building blocks that characterize the six possible movements between the bases as in this case where we describe the movement at the base step level. And here you can see the three translational movement, shear stretch and stagger and the three rotational buckle propeller and opening. And the same can be done at the base per step level where we can describe the geometry and the flexibility of each base pair by six movement, three translational or shift, slide and rise and three rotational movements, tilt, roll and twist. These descriptor are very useful to describe the overall geometry of the DNA and the macroscopic distortion of the DNA. And now I will show you an example of how the variation of one of these parameters can be connected directly to a biological example and now can be understood better why we are using this descriptor. So this descriptor is the roll angle, is one of the base per step parameters that explain one rotational movement. And here from the Kaladindru representation you can see that when we have a straight DNA if we apply the roll angle at two points we start to have a king structure of the DNA. But if we apply the roll like a periodic function we will have a bent DNA. Why is it important to be able to characterize the geometry of this movement? Because as you can see here is one of the main property of the DNA to be able to adapt the structure for the binding on the property. The roll is one of the parameters that is one of the most flexible and allows the DNA to adapt to this complex. Here for example we can see that DNA is highly bent and follow almost like this drawing bent conformation. And so the comparison between a straight DNA, an ideal DNA and the conformation in this complex will allow us to get more information about where their distortions are and also which step have been more distorted than the other. They probably are also connected with their ability to be flexed and to be distorted. Going back to the web server, this is the intergraphic interface that the user will face. So after uploading or after calculating trajectory using curves and flex will analyze the helical parameters. And what the user will be facing is the sequence, the first send for free prime to free prime and the other send for free prime to free prime on the other sense. So what the user will do is to select the base pair that is interested in to analyze the base pair and what we'll get is the values that can be the average results or the results by time. As I say before, for each analysis we will have the time course and the average results. For the average results, as you can see for each base pair, here are plotted the results for the user. Here you will see the role user and the average. And these values are not alone but as I say before have been compared with data base of DNA equilibrium geometries that is ABC and from the X-ray average. So the user could see if the DNA that have been studied is closer, for example, to the equilibrium canonical BDNA or is as some places where the DNA has more distortion and how it's closer or far away to the X-ray averages. The other way to visualize the data is as I say for each base pair parameters is along the time. This is the time assumption, so the time of the trajectory. Here is the value of role so I could see the user can see if there is a big variation or if it's a stable step and all the results summarize here. This is the distribution of all the values along the simulation and the user could see that mainly you will see the Gaussian distribution because the base pair parameters have usually a normal distribution and this is very important because in some cases it will be easier with using this graphical interface to analyze and to spot the base pair that has binormality or have two different minimals or have different conformations that are equally probable. Because the base pair parameters have a Gaussian distribution, harmonic distribution along the time, it's possible to analyze and to extract from this distribution the stiffness constant. I won't go into detail, if you have any questions you will see in the help session of the nflex server or you can ask me later or in the BioXcel forum. But mainly the stiffness constant are the constant that describe the stiffness of each base pair for the six different movement. And is described by a matrix of 36 elements, 36 constant is a diagonal matrix with pure term on the diagonal and mixed term between the different parameters. And these matrix can be calculated by the inversion of the covariant matrix obtained from either from the analysis of the MGT directory. So from the difference along the time of the MGT directory, it is the constant of Boltzmann, this is the temperature and this is the covariance matrix. But mainly what we're focusing on are the diagonal, so the pure term and they describe the flexibility of each step along the sequence. Why it is very important to know the stiffness constant. So here is the representation that you will find in the web page and in this matrix that is 36 elements and 36 values is a diagonal matrix. And in darker red is shown the constant that has the highest value, so the stiffest value. And we have to remind that translational and rotational movement have different units and degrees, so they have different order. And is important also to plot anything that is the main reason why the stiffness is important is to compare each base pair. For example, here is for the base pair parameter shift along the sequence, why? Because for the user will be easier to spot, mainly for an experimentalist, will be easier to spot flexible step that will be easier to distort if will be to form for example a complex or will be the part in which the motion of the DNA will go through. As has been shown before for the base step parameters, also for the stiffness, we have the values calculated by the user in red, the parameters that is the stiffness parameter for naked DNA in green and in blue for CHAMSO using another two different DNA phosphate. This comparison is very important, first of all, to see if the user is doing something wrong in case he is repeating a study, or if to show like maybe the sequence, the environment or the condition of the particular trajectory that is running the user are different, so the stiffness of the DNA change and this is important to focus on these changes. I will show you now an example on how in our group we use the knowledge of the stiffness to predict and compare experimental results. This work is published in the article DNA Structure Direct Expositioning and mitochondrial genome packaging protein that has been published this year. Here is an example of the structure of the DNA that is bound to a particular protein that is the ABF2P. The complex is formed by two main boxes, the MTG-1 and MTG-2, and these two boxes have an insertion you can see in blue, the residue that insert into the DNA and the KING DNA. This insertion provokes on the DNA a high end here, you can see on the right in the plot, an increase of role compared to the naked and bound structure. How did we use the flexibility for this complex? We predicted running just the unbound DNA that the sequence, the base pair with the high flexibility, so easily to distort with low cost stiffness constant will have been the TAAT where you can see here close the green arrow or yellow arrow. We predicted that this would have been the site of insertion of the base pair, while the stretch of polyetract followed by a T-tract that is very famous also in nuclear positioning to be a very rigid stretch. We predicted that it would have been in the region between the H&G box where the DNA is mainly straight. And as you can see, our prediction has been confirmed by the X-ray crystal structure. So this is a nice study to show how the flexibility and equilibrium parameter of the DNA can be used to predict and to confirm and to sometimes explain why we have some positioning of the DNA with the binding of a protein and why this happens. Now I will go through another tool that is another tool in NFLEX that is used to calculate the principal component of the system. So the principal component analysis allows to reveal the most important motion and can be applied mainly to protein and DNA. I want just to say briefly that as I say into the NFLEX server, NFLEX server can be run a simulation of the naked DNA, so unbounded DNA or also with protein. In case that the simulation involves also the protein, it's possible to calculate the principal components also for the protein, not only for the DNA, to see the complementary motion of both those molecules. This principal component analysis is basically a linear transform that extracts the principal vector of motion from the trajectory. In particular in this case is an essential dynamics of the DNA since with the principal component analysis, the essential motion of the DNA are extracted from a sample of a conformation. This sample or conformation are mainly the conformation along the trajectory. Here you will see a GIF, so you will see a representation of the motion of the DNA. This is very important first of all because many times the DNA is sort of like a static molecule or straight molecule and as you could see the molecule as a motion and in a particular direction. Having motion in a particular direction could explain sometimes why the protein is binding in one position or another or why the DNA is easily able to deform into one conformation or another. Going back to the NFLEX webpage here you will have the possibility to select the animation mode. Each mode is associated with eigenvalues and eigenvectors and only the first 10 that describe the main motions of the principal components can be selected and the user could see the movement of the DNA along the trajectory, the main motions. Now after describing the DNA at the base per level and the overall motion using the principal component analysis I will go through the microscopic descriptor in NFLEX at atomic level. I will start from the nuclear magnetic resonance observable. NFLEX allows the user to compare the results of the trajectory of the structure of the DNA or RNA to NMR studies. How nucleic acid, nucleic magnetic resonance observable that are the J coupling that here is a description that you will find also in NFLEX help part are scalar couplings between proton located three months away so can be described also by a diadral angle through the so-called Carpalus equation so can correlate the trajectory and the conformation along the trajectory and the experimental data using NMR and also the NOEs. The NOEs is the nucleic overhouser effect is a transfer of magnetization from one nuclear spin to another via cross relaxation and the proton-proton distance derived by NOEs are most useful for NMR parameters for structure elucidation so being able to extract using NFLEX server, NFLEX tools, the J couplings and NOE from the trajectory can allow the user to compare directly with NMR observable. Here is the graphical interface that you will see, here is the possibility to select the proton-proton position so between the sugar, between the sugar at the base, here is the sugar of the DNA and between the sugar and the base step. And also is important that here you can select the hydrogen distances for NOEs and for J couplings for the diadral that you are interested and so the user can check each parameter, the NOEs or the J couplings, compare them with NMR structure and this has been very useful for checking sometimes NMR structure but mainly to refine force field, why? Because sometimes building a new phosphate that is describing the DNA is important to compare with NMR or with experimental structure and the ability to transfer from the trajectory directly to NMR observables allows to refine and to see where the trajectory is failing sometimes or has some uncertainty about the results. Now I will go through other atomic analysis that is the canonical hydrogen bond analysis of the DNA, the distances and the backbone analysis. This for people that is not very familiar with DNA is the main pairing, is the Watson Creek pairing and are the hydrogen bond between two bases of the DNA and NFLEX allows to monitor along the trajectory or the average distance of the base pair canonical hydrogen bond. We found it very useful for example this analysis where we are analyzing complexes where the pairing is broken. So here you can see X-ray crystal structure, this is the PDB ID if you are interested and this is the un-pairing of a base that is called flipping out of a base so out from the double strand and usually it is caused by a protein you can see in purple the image of the protein. It is very, it is to gain more knowledge on the breaking of the base pair is interesting to run and dissimulation on the complex and monitor along the time how the base pair can get closer or far away and how for example the protein is acting to open for the flip out and the flipping of the base pair step. So this is one case in which this analysis of the breaking so the distance of the hydrogen bond is very interesting and has a very important biological impact. But the user can also decide not only to monitor the distance of these particular atoms that are involved in the canonical hydrogen bonds in the base pair but is allowed also to decide with atom pair distance to monitor and this is from the page of the web server so here the user could introduce the name of the atom one, atom two and check the distances and for example how along the simulation some atom will get far apart or some atoms will get closer and how the confirmation is changing for some during the dissimulation. Apart of the distances what can be analyzed as the hydrogen bond and stacking energy so we have the scripture at the atomic level not only of the distance so the geometry but also of the energy. What can be analyzed as the hydrogen bond stacking energies so the average, the minimum, the maximum and the standard deviation so how much they vary it along the time these are the four possibilities. One case where we found very useful to calculate the hydrogen bond and the pi stacking so the stacking between the base pair as you can see here or the hydrogen bond are the same one that I show you before has been for the airpin. Airpin is a very particular complicated structure is an example of X-ray crystal structure and the airpin is important to analyze the energetic and the distance because having a very, very significant stacking as you could see in many hydrogen bonds many times is important during the trajectory to evaluate which energetic component is the one that is stabilizing the complex and for example I have been studied and we are also simulation using our newly refined force field and we'll see that many times is the stacking energy that is stabilizing the complex more than the hydrogen bond but clearly this change along the MD simulation and is important to analyze how these two complementary forces are involved to the stabilization of the complex. The last descriptor, so the last analysis that can be done on the DNA here you can see the double strand as I say before this is the backbone that is composed by the phosphate and the sugar and these are the bases of the DNA and the backbone can be described by six main torsion angles around the covalent bonds and by the sugar packering. Here you can see all the descriptors, so the sugar packering here is the conformation of the sugar, here in north, here in south and mainly for conformation are dominating the conformation space. Here are the other angles, the five angles and the canonical for example alpha and gamma torsion that have been important in the formation of several proteins in a complex for example when we are analyzing a complex could be interesting to see how these two angles are varieting on how the protein is affecting these angles not only on the backbone, so not only on the conformation at the base pair level as we have seen before. And also we can analyze the B1, B2 population are experimentally known to exist at 80-20 so it's interesting to see how these two conformation can change in the particular case of a user. We use these tools to analyze the backbone angle and I will show an example to test a long time scale MD simulation and to compare the results of the geometry of the DNA at different environmental conditions. Here is the work that has been done by our group in particular by Pablo Danz and here they did the analysis of a long time scale molecular dynamic simulation of Dudrikerson-Dodekamer. Dudrikerson-Dodekamer is a DNA prototypical BDNA that have been tested experimentally many times so there are a lot of experimental data available and it's one of the main chain of DNA that can be used to test and to validate for example force field or how the DNA can behave in different ionic environment. In this case we tested the newly parameterized force field ParamBCC1 and in a variety of ionic environment and here for example you can see the time evolution of epsilon and zeta to be for the different base pairs also on the C3 to the G10 and this is the time evolution and we did the study to be sure that the DNA was exploring the canonical BDNA subset using different environment and we have seen also that only the cytonins and the guanins have more propensity in the transition between B1 and B2. Here you can see also for example for the same study I don't know if it's very visible. Here we monitor the six base pair parameters so shift, slide and rise up and then the upper panel and the bottom panel tilt, roll and twist. We monitor the base pair parameters of D to the camera using different ion environment so with potassium, with sodium a different concentration so only neutralizing the DNA our physiological concentration 0.15 molar or at high ionic concentration. So here we studied how the DNA can behave can for example increase or lower the stiffness or can open some base pair or the position of the ions can influence the geometry at the base pair level for all the simulation always using nflex. The last case that I'm going to show you is how we use nflex to compare the protein bound versus the unbound DNA conformation. Here is the structure that I've shown you before to see how the DNA is highly bent into this protein DNA complex and here for example for the parameter role we correlated the role for the naked unbound DNA so you can see in average the red values for each base pair for the naked DNA with the standard deviation in pink and in blue the role value for the DNA in the protein complex. So as you can see there are for example some steps where the protein is just acting pushing opening a bit the role values but the role is already high values in the naked DNA but here we can see a big difference where the protein is acting strongly on the DNA lowering the role. We also noticed that this big change is happening at the TA step that is usually characterized by low stiffness constant so high flexibility so it's easier for the protein to act on this step and to close the role for this step. So it's nice to compare how there are some ability of the DNA so properties of the DNA that are intrinsic to bend and to form the protein DNA complex and places where the protein is acting to form the complex. We also studied for this complex the variation for each base pair of the backbone angle here you can see epsilon and here you can see how in red the unbound DNA and the conformation of the DNA in the X-ray crystal structure and how the X-ray crystal structure is inside the standard deviation of the unbound DNA so in this case the protein is acting directly at the base pair level but it's not affecting at all the angle of the backbone of the structure. This is a summary of all the software that are involved in the part of NFLEX and for example we have CURS that is to analyze the base pair parameters there is PCI suite that is the program that has been used to analyze the principal component analysis T-LIP that is part of AMBER 2 for the setting up of the system for the trajectory and all these NFLEX can be also found in a comprehensive platform that includes a database system and an analysis portal for handling nucleic acid simulation so in this web server interface you could find MD simulation of a variety of unbound and protein bound DNA like for example the geodocuss under the camera that I just show you in different environmental conditions like the hairpin that I show you before and all the simulation are accessible and above all here is a screenshot of the Bignaz CMA web server interface and for each simulation NFLEX tools have been applied so we have available all the data all the analysis of NFLEX finally the NFLEX is also one of the analysis tools in a project providing tools and infrastructure to integrate the navigation in genomic data from the nucleic acid sequence to the chromatin level so in the multi-scale complex genomics project and in particular this virtual resource environment is possible to have tools to analyze DNA and RNA but mainly DNA from the atomistic level till the genomics scale these are all the tools that are integrated in the VRI of MUG and here for example you can see that there is NFLEX for the analysis of the DNA at the base parent atomic level and the web to be able to run simulation at atomic level of DNA strand or going closer to the genomic scale there are conscript models to analyze the nucleosome dynamics till tools and software to analyze experimental data at genomic level I would like to acknowledge and to thank my group and the professor Modesto Roscoe and Giuseppe Luigiep before starting this big project and in particular Adam Svitalla for that is the person that is developing these servers that allow the user to get closer to MD simulation and to the analysis tool for nucleic acids and if you have any questions now I know that there will be time for questions but if you have specific questions or you want to talk with us directly you can go in the bioexcel forum here is the web page and at the molecular dynamics you can see for example the link to the talk that I am giving just right now here is the bioexcel webinar given by for example by Adam Svitalla on the MD web as I told you before and if you have any questions you can look in and you can contact us directly so thank you for the attention and if you have any questions thank you very much indeed it's a very clear talk but I hope we will still have a question or two from the floor I know that some people arrived after I described how to do this so just to recap if you have any questions then you can type them into the question box in go to webinar so if you type your question there then I can either open the microphone and ask your question directly to Federica or I can read out your question so that you can answer it so if anybody has any questions you can type them into the box now while people are doing that or while they are thinking of anything they might want to ask or comment on then I will ask one quick question to me that one of the main advantages of this is that it makes it much clearer for novice users as to how to get access to all of these tools and how to step through these steps you also showed a list of all the different pieces of software that have been put together to create this which is an impressive list is the source code for any flex itself available if people are interested in seeing behind the scenes how these things are working or is that not the case at the moment for what I know I am not the developer of the web page of the web server but for what I know the source code is not available all the programs that you will see in the web page are easily downloadable and it is just a way to facilitate the analysis because the user doesn't have to download and can work directly using our web page but there is not an available source code for the web page but any question that the people can have for example on how all these programs are connected together can be asked in the forum or can be asked to us directly and we can provide any information I was thinking that question really more before I saw the slide where you showed all the software that was involved because I only asked the question because in the past we had some scientists who feel less ready to use a web portal because they are less sure about what code is actually running when they are doing something what method is actually being used but since all the code is available as open source I think that is quite obvious Yeah, all the program open source a person can just compare the results for example if they are not sure they can compare the results downloading the software is easy for example to visualize and to use them directly from the server then to download them directly there are not steps like there is a direct calculation and direct plot and also all the raw data that are calculated can be downloaded so the person can see directly if they are plotted correctly and they can plot by their own so there are not many filters in the program to the visualizations Yeah, that is good to know Thank you So, as you can probably tell I am not a direct expert in the scientific application area that Federica is talking about here so I don't have any particularly detailed questions but I was wondering if anyone else from the floor would like to ask you to ask a question Can we have any more just now So feel free while I am wrapping up to sneak in the last minute question if you want but otherwise we are getting close to the top of the hour now and as Federica has pointed out there are plenty of options for you if a question occurs to you later on so please do keep in touch with us if there is anything else that you would like to ask about but if we have no other questions from the floor I would like to thank you all for coming along this afternoon and I hope you will keep in touch with what BioExcel is doing and you can catch up again with the video when it is posted online in a few days time so thank you all very much for coming along this afternoon thank you Federica for a very interesting talk and yes everyone please do keep in touch with BioExcel thank you