 Okay, I think we can start now. First presenter is Dr. Nadezhda Todorova. She's a researcher at the Institute of Biodiversity and Ecosystem Research at the Bulgarian Academy of Science. She's interested in microbial ecology, implementation of molecular techniques for the assessment of ecological status, ecology and biomemetering of model prokaryotic systems, and others. She will present to us a talk with the Titan approaches for finding ligands inhibiting the NSP-10, NSP-16 complex of the SARS-CoV-2. So, you can start now. Thank you. First of all, boring dear colleagues, we present a part of our work. We are a big collective, and we are focusing on some research of finding inhibitors of one protein complex, which is part of the SARS-CoV-2 protein. The problem is well known, so the number of global deaths is constantly rising around the globe. This is a new virus that has some dangerous features, which are important for its spread, namely the infected person is contiguous while being long asymptomatic. And there is no cure available connected to the high pressure on a medical system. It leads to a high mortality rate. It's a global scientific community is mobilized since the beginning of this global pandemic in a search for appropriate drug. And just a minor note, which I think is really unfortunate, searching the Johns Hopkins Institute database, found that now Bulgaria has the highest mortality and second fatality rate in the world. Our target on the right-hand side is presented major steps of viral vital cycle. And first, we had to identify targets for the antiviral. We needed the conservative part of the virus. We needed part of it that is crucial for the cycle of reproduction. And it should not be shared with the whole cell, I mean not the one part of the cell own machinery. We focused our interest on an enzyme composed of two parts. This is non-structural protein 10 and non-structural protein 16, that together form the two-prime hydroxy-metallase complex. What is it important for? It uses a cell molecule. This is the S adenosyl methionine, a molecule that the cell by its own uses for making a specific methylation of the RNA. This is making a cap. And this helps the virus to mimic the own cell molecules and thus invade the cell, muting the cell defense mechanisms. On the left-hand side are presented the starting molecules of the enzyme molecule, which is the ligand of this protein complex and the end product, mainly taking away the methyl group. What was our plan? We are trying to find the ligand that is capable to captivate this viral proteins by blocking the sun pocket. It was ambitious in two ways. First of all, we had to find the ligand that can bind the pocket to the existing folks of drugs, drugs, candidates, metallurys, that is enormous in quantity. And the second was that we needed to evaluate the binding energy and in a way, good enough to distinguish among the best ligands. To fulfill this task, we should first extract maximum information that is available as this is a new target. Second, we needed to define the best possible working model of the protein and its sun pocket to unify the optimal model for scoring of ligand and protein interaction to use and find optimal docking procedure for achieving the existing pool of structures. There are several methods for generating also new ligands, hoping some of them will be better. And to calculate adequate binding energies of ligands is using the molecular dynamics. So emphasis placed on relatively simple for synthesis compounds with little or no asymmetric carbon atoms or which asymmetric carbon atoms in symptoms are easily available as a synthetic equivalent. We were lucky as there were lots of published crystallized structure of our protein in the pocket with the appropriate ligand, not only for the SARS-2, but also for the previous studied, more studied viruses in this group, namely the SARS and MERS. Here is the list of structures which we used and the max two RDS which took our attention, so with which we worked further due to their good resolution and minimal amount of missing parts. The missing parts were cured using the code software, other minor conformational problems were resolved using the capabilities of the MOS software. Protonation of the proteins is done also with the tool from the MOS software, the protonate 3D application form and the listed parameters are also included in the down part of the slide. On the right side is presented what we found, namely the geometry of the pocket is highly conservative, which was really great. Not only in the group of crystallized structures of COVID-2, SARS-2, but also the other two. To define a reliable model of our pocket, some pocket in protein complex we needed a properly hydrated and equilibrated structure as the physiological conditions are completely different compared to the crystal one. So we need MD simulation. We have tried different software for this task. The biggest problem was the parameterization of the sub molecule because there was a lack of parameters for the unique sulfonium ion. We performed MD simulations with different software and different force fields. They are listed and we proceeded further with the MD trajectory obtained with the SARS because we planned to use amber in docking and binding energy calculations. And based on these and as a parameterization of ligands with the SARS, it's important for binding energy calculation by molecular dynamics. The depicted picture on the left protein in the complex is the NSP10 in blue. On the right side is the NSP16 in green. It's visible. The pocket for the sum is visible. There is the sum itself. In dark, blue and green are the crystal structures. The light, the colors are what is visible at the end of the MD simulation. The last two turrets from the NSP some equilibrated trajectory was used further for the cluster analysis with chromax. The most populated cluster was used. It comprises 66%. And the frame with the lowest RMST to the central structure of this named cluster one was used further as a pocket template. Closed amino acids, amino acid residues that were close to the sum were minimized at unbar-enhanced Huckel theory force field with R-field solvation model as it is implemented in the MOS software and this method is used in further steps. Because we're using use fit strategy in the fit and scoring of the ligands we needed to optimize the pocket in the close vicinity of the sum as the selected frame from the molecular dynamics not all interactions are optimal. And the pictures below the left side is presented a pocket with the sum in it and the interacting amino acids on the right-hand side the same thing but after the mentioned procedure of minimization. What they could be seen is that the sulfur atom along with it's both metal groups as well as two prime and three prime hydroxy groups from the ribose ring are not buried in the pocket but rather exposed to the solvent. The pocket itself here is based on two methods used to assess it. In green are the lipophilic parts the hydrogen bonding parts are in capable of hydrogen bonding parts are in purple and in blue is the minor polar part and in the center is the sum. So what is important is that the pocket is quite regular but it has two pits. The one is adopting the amino acid part and it's wider while the other which is adopting the adenine residue is quite narrow and there's only aromatic residues can reside inside without van der Waals repulsion. This is a statement that we further approved and information on it will be further in the presentation. Main SAM and SIH interactions with the amino acids of the pocket are listed on the right-hand side of the flight. The most strong interaction was with the amino group from the amino acid part followed by the amino group of the adenine part and the carboxyl group of the amino acid part too. And what we found also interesting is that the SAM interacted better with the pockets of the selected frame. We further made a confirmation on this of the pocket in order to clarify the mobility it was made with and without SAM. So we performed two conformational tasks. On the picture here is presented the result of the protein part with the global with the SAM molecule in its global minima and in local. The interesting was that the global potential minima of the pocket with SAM is about 21 kilocalories per mole lower than the first stable local minima. And as it can be seen on the picture in red is the SAM in global minima while the other colors of green, blue, yellow are the same molecule but in local minima. We've seen that only the amino acid part is right in its pocket. So we found during this confirmation analysis more information about the pocket and the exact placing position of the ligand in it. The major contribution in the overall SAM pocket interaction energy is based on the amino acid part. And in complexes with lower or modest interaction energy this was the only part that was placed in the right place. And these conformations could be used to trace the entrance of the SAM in the pocket. From interaction maps of this conformation could be seen that we have some guiding parts of the pocket. So amino acids that are not so close to the ligand but are crucial for positioning in the right position. There were two as part of the acid residues that guided the ribose part and one as per again that was helpful for placing at an important place. The result from the research of the pocket without some molecule showed that only three kilocalories per mole from the global minima there are wide differences in positions of amino acids that are much larger than the other. And these positions of amino acids that are marked as essential for the SAM pocket interaction. And such mobility of the empty SAM pocket requires usage of induced feed procedure for optimal arrangement of pocket toward ligand. This could be seen on the picture the ligand so the same molecule is placed just as a reference after the end of the conformation changes. We followed with multi-fragment search to understand the interactions between chemical functional groups fragments with the active site of the receptor. We placed a number of such functional groups they are listed in the last part of this slide. The exact procedure is also described so we put a number of fragments 300 copies in our case and every in a random manner and everyone can see only the not the other copies but the pocket and the pocket or the protein itself can see all of them. In our case we used this approach to clarify the most potent amino acids of the pocket interaction which chemical functional groups are best interacted with it and also what is the amount of the energy that can be achieved from such interaction. It is presented one of these multi-fragment search namely this one in which was used phenol and these are both pits of the pocket can be seen on the right side where is the adenine part it is quite narrow so the ligands must have an aromatic ring. Other lipophilic compounds that were tested were too big even the propane cannot enter that pocket. So this will be used in the primary selection criteria further for our set of ligands. Analysis of the results obtained from the multi-fragment analysis are listed in this slide so it should have aromatic ring it is necessary to have two primary or better secondary amino groups in order to interact with the guiding amino acid residues of the pocket due to the geometry of the pocket between these two amino groups should be between 8 and 18 atoms and one of the amino groups should be connected to aromatic ring. These criteria were used for primary selection of the ligands for further screening of the variables of the ligands. An important part of methodology we need to optimize and unify ligands for the ligand pocket optimization methodology and we need it optimal and unified method for binding energy calculations in order to compare binding of ligands obtained from differences. For scoring we used two scoring functions London Delta G and GBVI YSA Delta G scoring function to estimate the free energy of binding of the ligand from a given post because the first method was not a resource demanding it was used as a first step so something like a saving step but further optimizations and were you all ligand were estimated using our unified GBVI YSA Delta G force field function so we made a steep approach to gradually see best ligand from the initial set and the correlation between the two functions is not linear we tested minimum 25% of the pool that was stated best using our pre-estimation methods the London Delta G and after that once again calculated using the unified ones important words on the ligand placement this is really important step we tested the exact placement technology of the ligand in the pocket is crucial even the best optimization of the ligand inside the pocket can't help in a badly placed ligand we checked several different placement strategies the procedure is described here in this point and from the four different placement methods in the case with the placement according to the pharmacophore model moreover the pharmacophore model was also used as a filtering matching the poses to the pocket we should think so making our pharmacophore we needed to stay wider in the spheres to be used in this model more possible positions of functional groups because of the mobility of the pocket the groups are listed here but we also needed to make the wireness of the spheres not too big to be possible to guide the sub-molecule when pharmacophores are used as docking they are done ambiguously to the position in the pocket positioning of the ligands in the pocket in some local global minima is then relatively rough during the first stage of the screening and placement of the ligands because of the mobility of the pocket residues in that stage we used for the large subsets the gritmin refinement in the case the refinement was done with Huckel-Telre force field in the last stage of the induced fit scoring to measure interaction energy capable to order poses for the stage of MD simulation we need a precise and reproducible induced fit positioning of the ligand what we found best from the parameters listed here chosen method some binds to the pocket with minus 12.7 kilocalories for most of this value was used in the further work as a trace for food I'm just sorry to have to put you have to kind of orient towards the summary of your result story I'm not sure I understood you have two or three more minutes to finish, yes okay I thought I have half an hour after the start, okay I just wanted to leave time for the questions, okay okay I'll try far away from the end but I'll go through the slides quicker the presentation of how it looks like the induced fit which movable and not movable but seen electrostatically I mean as it's from the pocket docking was done commercially available set up ligands and here are listed the databases the problem with the database with that components inside are in smiles that makes the crucial problem is the placement of the ligand in their conformations the placement mechanism uses only torsion driver and it can generate adequate conformation so we used the conformation in Porter in Des Moines the Zinc 15 database on the right side is the algorithm which we used and what we have chosen from the database on the left side with database with drug structures for further docking at the end mentioned the enaminerial which is enormous so more than one billion but relatively simple and easy for synthesis compound that were achieved using our criteria the drugman's search using PostgreSQL with the bingo cartridge from APAM Vuxy database having novel drug like scaffolds is also tested and can be manually curated database last approach that was the ligand hoping and three methods came scaffold replacement here just to mention that we had three parts of the molecule which we used in the scaffold replacement in three different sets but we haven't changed parts with the arrow mentioned here that are essential for the interaction of the complex so these are the amino groups and the carboxyl that were analyzed in the beginning was important all of them were subject to the induce feed to rescoring and ordered by their interaction using our unified energy scoring function breeding was made based on the full combination of previous ligands that were resultant from the scaffold replacement giving a generation of yet close to one million new valuable structures all of them again we scored by induce feed procedure and current results our approach resulted in more than 17,000 compounds better than some on the upper right corner is presented the heat you know which had 14% better binding energy than the some molecule the interaction energies of the best hits that passed admit currently are calculating on the MD level and we hope we have the possibility to test the best hits from them experimentally at the end I would like to thank to say thank you for the support to praise to national science from the Bulgarian national center for supercomputing application and the Institute of information and commercial technologies thank you for your attention thank you so I saw that we have one question in the chat