New MurD synthetic targets

[danaklug]

Based on @KatoLeonard's analysis of the first round MurD activity assay results, we're working on some follow up compounds.

I'm focusing on the expanding the piperazine/isopropylamine side, which wasn't explored much in the first set of compounds. Based on the binding poses of the original hits below, it seems like there is room to fill some space by trying to expand into the highlighted areas. However I'm certainly not a modeling expert and am happy to get second opinions on this analysis or thoughts on whether attempting to dock some proposed compounds into this pocket would be productive.

The synthesis, structures, and SMILES of my proposed analogues are below:

CC1=CC=C(NC(C2=CC(NC(C)C)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(NC(C)(C)C#C)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(NC3CC3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(NC3CCCCC3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(NC3CCNCC3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(NC3CCN(C(OC(C)(C)C)=O)CC3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(NC3=CC=CC=C3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(NC3=NC=CC=C3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(NCC3=CC=CC=C3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(N3CCCC3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(N3CCCCC3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(N3CCOCC3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(N3CCS(CC3)(=O)=O)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(N3CCNCC3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(N3CCN(C)CC3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(N3CCN(C(OC(C)(C)C)=O)CC3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(N3CC(C=CC=C4)=C4CC3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(N3CCN(C4=CC=CC=N4)CC3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(N3CCN(C4=CC=C(OC)C=C4)CC3)=CC=C2)=O)C=C1 CC1=CC=C(NC(C2=CC(N3CCN(C4=CC=NC=C4)CC3)=CC=C2)=O)C=C1

[KatoLeonard]

While Dana works on expanding the solvent exposed side of the fragments, I focus on the phenyl side embedded in the pocket. I have constructed a library of substituted phenyls that, in my opinion, make sense based on the size, shape and residues of the pocket, in combination with the SAR diagram.

This is the original fragment hit 373 bound to murD. It seems that the pocket isn't that large, so not much space for expansion of the fragment on this side. That's why I tried to build in essential H-bond donors/acceptors that could possibly form new interactions with residues in the pocket.

Library structures, synthesis and SMILEs:

O=C(NC1=CC=C(C2=C(N)C=CC=C2)C=C1)C3CCN(CC3)C O=C(NC1=CC=C(C2=CC(N)=CC=C2)C=C1)C3CCN(CC3)C O=C(NC1=CC=C(C2=CC(CO)=CC=C2)C=C1)C3CCN(CC3)C O=C(NC1=CC=C(C2=CC(C(N)=O)=CC=C2)C=C1)C3CCN(CC3)C O=C(NC1=CC=C(C2=C(F)C(F)=CC(N)=C2)C=C1)C3CCN(CC3)C O=C(NC1=CC=C(C2=CC(N=CN3)=C3C=C2)C=C1)C4CCN(CC4)C O=C(NC1=CC=C(C2=CC(OC=C3)=C3C=C2)C=C1)C4CCN(CC4)C O=C(NC1=CC=C(C2=CC(OCO3)=C3C=C2)C=C1)C4CCN(CC4)C O=C(NC1=CC=C(C2=CC(NC(N3)=O)=C3C=C2)C=C1)C4CCN(CC4)C O=C(NC1=CC=C(C2=C(C(OC)=O)C=CC=C2)C=C1)C3CCN(CC3)C O=C(NC1=CC=C(C2=CC(OC)=C(OC)C=C2)C=C1)C3CCN(CC3)C

I tried modelling some of them, but have no experience in the field whatsoever. Would greatly appreciate some feedback on the structures and keen to see results of cross-docking of my structures.

[ZigBu]

@KatoLeonard

Hi Kato： Thank you so much for the exquisite model schematic diagram and clear explanation. Really appreciate that!

However since I'm not an expert in organic chemistry either, I may have a question about your library structures and SMILEs: The original fragment hit 373 has a tertiary amine beside the amide group. Would you please explain the rational behind changing the N atom into C atom if it's not too much trouble. Also would you mind provide the information about which MurD enzyme did you use as the target. Thank you so much!

Many thanks and wish you all the best! Zige

[KatoLeonard]

Hi @ZigBu,

Good question! So before I joined the project, Dana had already done a round of elaboration on the fragment hits. Her rationale was that by changing the piperazine to a piperidine, the compounds would be easier to synthesise. There were two reasons why it would be appropriate to change the indicated nitrogen of fragment 373 to a carbon. First, this nitrogen does not interact with the target, so there was no risk of losing essential interactions. Secondly, the top nitrogen is protonated and makes an H-bond with glutamic acid. In a piperidine, this nitrogen should be more basic than in a piperazine, which means a stronger interaction. The 'round 1' elaborated fragment library was tested for inhibition activity against murD and some piperidine-containing fragments were active, confirming the assumptions. Hopes this answers your question!

Best wishes, Kato

[ZigBu]

Hi dear @KatoLeonard : Thank you so much for you clear explanation! Really appreciate that! I'm sorry to bother you again！ Would you mine provide the information about which strain of MurD are you targeting and would you ming to provide the PDB code for the enzyme. That would be really helpful!

Many thanks best wish！ Zige

[MFernflower]

@KatoLeonard Hello - I don't currently have access to a docking pipeline but was curious about your opinion of this idea I had:

[KatoLeonard]

Sorry @ZigBu, I forgot to answer your second question! The pdb code is 3LK7 and it's the crystal structure of UDP-N-acetylmuramoylalanine-D-glutamate (MurD) ligase from Streptococcus agalactiae. You can access the pdb file of all the original fragment hits bound to murD here. I think you have to change the .txt extension to .pdb yourself when you download the file. If there is anything else I can help with, just let me know!

[ZigBu]

According to the SeeSAR result, the compound1, 4, 5, 6, 8, 11 created by Kato have the potential for the increasing affinity. However, I am not familiar with the software yet, thus there may be some compounds missing. If someone could double check, it would be fantastic.

Molecular docking is done by Glide module in Schrodinger Maestro software (free trial). The results of binding energy (docking score) are shown in table1: (DOI)

The results show that, the best in docking scoring with MURD targets is Compound 4 and 11. According to these two 3D figures obtained by PyMol, my observation is that the compound can fit in to the target binding pocket. However there is some questions that will be discussed later.

FIGURE 1: The surface electrostatic surface of MURD with Compounds Left: electrostatic surface of MURD with Compound4 Right: electrostatic surface of MURD with Compound11 (prepared by PyMol)

Then the compounds with the highest binding score are selected and the interaction between ligand and residues is plotted by MOE (free trial) for further analyzing.

FIGURE 2: The surface electrostatic surface of MURD with Compounds Left: the detail binding mode of MURD with Compound4 Right: the detail binding mode of MURD with Compound11 (prepared by MOE free trial)

Since I am not an expert in molecular docking or organic chemistry, I have some questions about the analysis of the results.

Question 1: Wether the compound is able to fit target binding pocket? According to figure1 , both compounds can fit into the bind to the target binding pocket. However in the Figure two , there are fuzzy blue blob, which refer to the part of the atom of the ligand or the residue of the protein is exposed to the solvent. (DOI). So does that mean we can keep adding other functional groups here, or there is no room in this binding pocket?

Question 2: Why the binding energy is slightly low for compound11? Compared with compound4, there are three more side chain acceptors for compound11. Do these bonds contribute to the stabilization of small molecules with residue Glu132? If so, then why the binding site for that compound is slightly higher than compound4.

[ghostbio]

dear zige and all-

i have only very briefly followed things here, but compounds that show an indication of potential binding in SeeSAR, that has caught my attention.  :)

SeeSAR's affinity calculation is based on HYDE, and the visualization of the affinity is ONLY based on hyde. so, if you want to be more re-assured that there may indeed be something to it, then you must please co-watch torsional strain and clash indicators.

i am happy to organise a, say, 2h virtual workshop for everybody in the murligase network who wants to get started with SeeSAR. (= an intro/kick start)

please note that we have the YoungSolvers initiative for students working on their master theses (or the like). biosolveit.de/YoungSolvers is your friend.

mail me with "murligase" contained in the subject field and, depending on feedback, i will set something up.

tons of success and thanks you all, and, who knows… until soon!

  marcus

--

Dr. Marcus Gastreich                @.*** Sr. Director Application Science Phone: +49-2241-2525-0                           www.biosolveit.de BioSolveIT GmbH - An der Ziegelei 79 - 53757 St.Augustin - Germany facebook http://facebook.com/BioSolveIT                                            linkedin group http://www.biosolveit.de/linkedin

Zige Bu schrieb am 23.11.21 um 23:21:

According to the SeeSAR result, the compound1, 4, 5, 6, 8, 11 created by Kato have the potential for the increasing affinity. However, I am not familiar with the software yet, thus there may be some compounds missing. If someone could double check, it would be fantastic.

Molecular docking is done by Glide module in Schrodinger Maestro software (free trial). The results of binding energy (docking score) are shown in table1: (DOI https://pubs.acs.org/doi/10.1021/jm0306430) image https://user-images.githubusercontent.com/92034546/143138108-765cc7f4-504d-431a-8af5-18a8d83450c7.png

The results show that, the best in docking scoring with MURD targets is Compound 4 and 11. According to these two 3D figures obtained by PyMol, my observation is that the compound can fit in to the target binding pocket. However there is some questions that will be discussed later. 截屏2021-11-23 17 02 24 https://user-images.githubusercontent.com/92034546/143137562-58809f52-ccf4-4069-9bb5-870eecb127a4.png FIGURE 1: The surface electrostatic surface of MURD with Compounds Left: electrostatic surface of MURD with Compound4 Right: electrostatic surface of MURD with Compound11 (prepared by PyMol)

Then the compounds with the highest binding score are selected and the interaction between ligand and residues is plotted by MOE (free trial) for further analyzing. 截屏2021-11-23 17 03 56 https://user-images.githubusercontent.com/92034546/143138127-250cce0e-19c1-43e4-a2bb-efab574ab0ea.png FIGURE 2: The surface electrostatic surface of MURD with Compounds Left: the detail binding mode of MURD with Compound4 Right: the detail binding mode of MURD with Compound11 (prepared by MOE free trial)

Since I am not an expert in molecular docking or organic chemistry, I have some questions about the analysis of the results.

Question 1: Wether the compound is able to fit target binding pocket? According to figure1 , both compounds can fit into the bind to the target binding pocket. However in the Figure two , there are fuzzy blue blob, which refer to the part of the atom of the ligand or the residue of the protein is exposed to the solvent. (DOI https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0086455). So does that mean we can keep adding other functional groups here, or there is no room in this binding pocket?

Question 2: Why the binding energy is slightly low for compound11? Compared with compound4, there are three more side chain acceptors for compound11. Do these bonds contribute to the stabilization of small molecules with residue Glu132? If so, then why the binding site for that compound is slightly higher than compound4.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/opensourceantibiotics/murligase/issues/58#issuecomment-977223909, or unsubscribe https://github.com/notifications/unsubscribe-auth/AMA3GKFA3GHRZ7GE6REMMBDUNQHWXANCNFSM5IBNYMIQ. Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

[ZigBu]

@ghostbio Dear Marcus: Thank you very much for your reply. This is very timely for us. I still have a lot of confusion in terms of operation, since I am unable to find enough SeeSAR-related guidance or tutorial. At present, I can only use it as a basic docking reference. Specific docking and drawing can only be plotted with other commercial software, but they are all time-limited free trial versions.

We would be really appreciate it if we could have a 2h virtual workshop together. Many thanks. Wish you all the best!

[KatoLeonard]

@ZigBu thank you for all the work! Really interesting to see the results of docking done with other software than the Jupyter notebook on the CompChem Tools page. I was wondering how you analysed the different poses obtained by Glide, or did you just pick out the poses with the lowest binding energy? It seems that the pose of compound 4 you're showing is quite different from the pose of original fragment hit 373 obtained by X-ray screening, and I thought I heard that it is best to select poses that resemble those of the original fragment bound to the target, which is shown in the attachment. But I may be wrong about this. @Yuhang-CADD Can you confirm or deny that?

@ghostbio really keen to do a virtual introduction workshop to learn about SeeSAR. Thanks for willing to organise this!

[ZigBu]

@KatoLeonard Hi Kato, for the poses, I just selected the pose with the lowest binding energy. If it is best to select poses that resemble those of the original fragment bound to the target, then I will make the dock again! Since the software automatically select the poses with the best docking score, I did not look into other different poses. However according to the docking model, I was wondering that the reason why compound 4 is different from the pose of original fragment hit 373 might might due to the a stronger interaction between Nitrogen at the amide group and that surrounded water molecule. I am not sure does this contact may enhance the stability between compound or quite the opposite？

[ZigBu]

Update: By taking the instruction form Kato into consideration, I selected two compounds with the poses that resemble those of the original fragment bound to the target as shown below. While on the other hand, all the other compounds share a similar poses which point to an opposite direction compared with the original 373.

[ghostbio]

@ZigBu / @KatoLeonard-

for the SeeSAR workshop: what time zones are you in?

thanks! marcus

 

Kato Leonard schrieb am 24.11.21 um 11:42:

@ZigBu https://github.com/ZigBu thank you for all the work! Really interesting to see the results of docking done with other software than the Jupyter notebook on the CompChem Tools https://github.com/opensourceantibiotics/murligase/wiki/CompChem-Tools page. I was wondering how you analysed the different poses obtained by Glide, or did you just pick out the poses with the lowest binding energy? It seems that the pose of compound 4 you're showing is quite different from the pose of original fragment hit 373 obtained by X-ray screening, and I thought I heard that it is best to select poses that resemble those of the original fragment bound to the target, which is shown in the attachment. But I may be wrong about this. @Yuhang-CADD https://github.com/Yuhang-CADD Can you confirm or deny that? Screenshot 2021-11-24 at 10 41 44 https://user-images.githubusercontent.com/91056178/143223486-ba1d9685-cbb6-418c-b355-adadd9ca132a.png

@ghostbio https://github.com/ghostbio really keen to do a virtual introduction workshop to learn about SeeSAR. Thanks for willing to organise this!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/opensourceantibiotics/murligase/issues/58#issuecomment-977752462, or unsubscribe https://github.com/notifications/unsubscribe-auth/AMA3GKA3FWE7KI2BEQPM53TUNS6TBANCNFSM5IBNYMIQ. Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

--

Dr. Marcus Gastreich                @.*** Sr. Director Application Science Phone: +49-2241-2525-0 / Fax: -525               www.biosolveit.de BioSolveIT GmbH - An der Ziegelei 79 - 53757 St.Augustin - Germany facebook http://facebook.com/BioSolveIT                                            linkedin group http://www.biosolveit.de/linkedin

[ZigBu]

@ghostbio Thank you so much for asking. I am in London GMT now.

[Yuhang-CADD]

@KatoLeonard Hi Kato, for the poses, I just selected the pose with the lowest binding energy. If it is best to select poses that resemble those of the original fragment bound to the target, then I will make the dock again! Since the software automatically select the poses with the best docking score, I did not look into other different poses. However according to the docking model, I was wondering that the reason why compound 4 is different from the pose of original fragment hit 373 might might due to the a stronger interaction between Nitrogen at the amide group and that surrounded water molecule. I am not sure does this contact may enhance the stability between compound or quite the opposite？

Hey, nice question by @KatoLeonard I was wondering the same as Kato about how you (@ZigBu) selected molecules and how you managed to compare the poses with the original hit in the pocket (did you do it via visual inspection or by RMSD analysis?) Would you mind sharing your exact steps of doing the docking and pose selection (maybe a link to your ELN with details in words and screenshots)? Really appreciate it if you could, as we would find it a bit difficult to understand the process if not! Thanks!

Also, @ZigBu the programme would automatically select the pose because you forgot to set the amount of conformations to be generated, thus it was set as a default “1”. So it would be great if you could inspect into all of Kato’s molecules with poses similar to the origina hit. Normally I would prefer to analyse the pose similarities by checking both RMSD values and via visual selection.

Hope this helps! Many thanks, Yuhang

[ZigBu]

Thank you so much for your advices and guidances. @KatoLeonard @Yuhang-CADD As a beginner in this area, I did almost all the steps according to various video tutorials, so I can only describe each step in general. And some reference videos are explained in my native language, so it may be difficult for you to understand.

Selection part: To start with, the file contained the original hit of 373 fragment with 2LK7 is uploaded to the SeeSAR software and then the binding site, where the 373 is targeted, is selected automatically by the software(binding site mode). Then I upload all the compounds that is created by Kato to the docking mode of SeeSAR. Some compounds only have 1-2 poses that is able to bind to the target binding pocket and the remain compounds I selected have at least 3 poses that does not have the red mark shown by the SeeSAR(docking mode). These steps are done by following the video tutorial on YouTube and other Video web site. （DOI）（DOI）

However, as I just mentioned, I am not familiar with the software, and I only got the free trial version. Thus, the selection is only for reference, and I agree that it would be great if I could inspect into all of Kato’s molecules with poses like the original hit. As my original thought is that SeeSAR as a brand-new software might be helpful in selecting large number of compounds in the future? Schrodinger's Protein Preparation Wizard was used to process the Protein

Docking part: Prepare compound All the selected compounds were originally plotted by ChemDraw and then converted to 3D structure and prepared with minimized energy by Chem3D. (DOI) Then all the structures were imported into Schrodinger software to establish a database. All molecules were prepared according to the default Settings of LigPrep module. (DOI)

Prepare protein Schrodinger's Protein Preparation Wizard was used to process the protein, including remove water and minimize energy. （DOI）

Molecule docking Molecular docking is done by Glide module in Schrodinger Maestro software. During screening in Glide module, the prepared receptors are imported to specify appropriate locations in receptor mesh generation. Each step is revealed clearly in the video tutorial (DOI） (DOI)

During this process, I couldn’t find any tutorial to reveal different poses of the binding, thus I only output the binding with the lowest binding energy.

Comparison part: I apologies for not take poses into consideration, thus the program defaults to the poses with the lowest binding energy. Then the complex of protein and small molecule was visually analyzed by Pymol and the latest update reveals that the compound2 and 11 have the most similar binding poses. As for the analyzing of RMSD values, I would take an insight later.

I would love to inspect into all of Kato’s molecules with poses like the original hit. However, my license file for Maestro and ChemDraw are only 10 and 14 days respectively, and they are about to expire by this Friday. I am worried about this and want to know what is the solution? Since I have limited experience in scientific research, I am very sorry to bother you guys. Thank you so much for the generous help.

Many thanks! Wish you all the best!

[Yuhang-CADD]

Nice work Zige! I am impressed that you are really good at self training!

The process should be like that.

I think you can also try SMINA (introduced in our original docking tutorial paper) as it has rDkit installed to generate as many as 20 max conformations for each of Kato’s molecules.

Best, Yuhang

On Wed, 24 Nov 2021 at 17:43, Zige Bu @.***> wrote:

Thank you so much for your advices and guidances. @KatoLeonard https://github.com/KatoLeonard @Yuhang-CADD https://github.com/Yuhang-CADD As a beginner in this area, I did almost all the steps according to various video tutorials, so I can only describe each step in general. And some reference videos are explained in my native language, so it may be difficult for you to understand.

Selection part: To start with, the file contained the original hit of 373 fragment with 2LK7 is uploaded to the SeeSAR software and then the binding site, where the 373 is targeted, is selected automatically by the software(binding site mode). Then I upload all the compounds that is created by Kato to the docking mode of SeeSAR. Some compounds only have 1-2 poses that is able to bind to the target binding pocket and the remain compounds I selected have at least 3 poses that does not have the red mark shown by the SeeSAR(docking mode). These steps are done by following the video tutorial on YouTube and other Video web site. （DOI https://www.youtube.com/watch?v=TCnBviIVgXY）（DOI https://space.bilibili.com/360143874?spm_id_from=333.788.b_765f7570696e666f.1 ）

However, as I just mentioned, I am not familiar with the software, and I only got the free trial version. Thus, the selection is only for reference, and I agree that it would be great if I could inspect into all of Kato’s molecules with poses like the original hit. As my original thought is that SeeSAR as a brand-new software might be helpful in selecting large number of compounds in the future? Schrodinger's Protein Preparation Wizard was used to process the Protein

Docking part: Prepare compound All the selected compounds were originally plotted by ChemDraw and then converted to 3D structure and prepared with minimized energy by Chem3D. ( DOI https://www.bilibili.com/video/BV155411j7Mj?from=search&seid=3619328781678564092&spm_id_from=333.337.0.) Then all the structures were imported into Schrodinger software to establish a database. All molecules were prepared according to the default Settings of LigPrep module. (DOI https://www.bilibili.com/video/BV17p4y1h7mG?from=search&seid=9837401913949597352&spm_id_from=333.337.0.0 )

Prepare protein Schrodinger's Protein Preparation Wizard was used to process the protein, including remove water and minimize energy. （DOI https://www.bilibili.com/video/BV155411j7Mj?from=search&seid=3619328781678564092&spm_id_from=333.337.0.0 ）

Molecule docking Molecular docking is done by Glide module in Schrodinger Maestro software. During screening in Glide module, the prepared receptors are imported to specify appropriate locations in receptor mesh generation. Each step is revealed clearly in the video tutorial (DOI https://www.bilibili.com/video/BV18h411f7AC?from=search&seid=4786928550062325486&spm_id_from=333.337.0.0） (DOI https://www.youtube.com/watch?v=QBy_flzNAgw)

During this process, I couldn’t find any tutorial to reveal different poses of the binding, thus I only output the binding with the lowest binding energy.

Comparison part: I apologies for not take poses into consideration, thus the program defaults to the poses with the lowest binding energy. Then the complex of protein and small molecule was visually analyzed by Pymol and the latest update reveals that the compound2 and 11 have the most similar binding poses. As for the analyzing of RMSD values, I would take an insight later.

I would love to inspect into all of Kato’s molecules with poses like the original hit. However, my license file for Maestro and ChemDraw are only 10 and 14 days respectively, and they are about to expire by this Friday. I am worried about this and want to know what is the solution? Since I have limited experience in scientific research, I am very sorry to bother you guys. Thank you so much for the generous help.

Many thanks! Wish you all the best!

— You are receiving this because you were mentioned.

Reply to this email directly, view it on GitHub https://github.com/opensourceantibiotics/murligase/issues/58#issuecomment-978094552, or unsubscribe https://github.com/notifications/unsubscribe-auth/AN6TV5C5ZGHHB2KSSBCFWQDUNUP3RANCNFSM5IBNYMIQ .

[KatoLeonard]

Hi @ZigBu, thanks for explaining your process. Very intriguing, well done! As regards the license for ChemDraw, I installed it through the UCL software database: search for ChemBio Office Ultra > Licenses > Personal Computer use information: click here. On the PerkinElmer site search for UCL and register for ChemDraw. In terms of the Maestro license, I don't remember how I got access to it since @Yuhang-CADD helped me. Maybe he can explain it?

@ghostbio same as Zige, located in London, GMT time. Many thanks!

[ghostbio]

dear all-

several of you said they'd appreciate an introduction into SeeSAR.

i am happy to set this up, here comes a doodle survey for a 2h workshop:

   https://doodle.com/poll/xikncwzaffcxcztx?utm_source=poll&utm_medium=link

please note that all these proposals are BERLIN time. (most of you seem to be in europe as it seems —??)

for the workshop we will equip you with a 6 weeks license as a jump start, so once i have the time slot, i will please need your email addresses.

please note: longer term licensing for 0$ for academics is possible through our 2 initiatives:

 https://www.biosolveit.de/YoungSolvers  and  https://www.biosolveit.de/scientific-challenge/

SeeSAR can be downloaded from biosolveit.de/download

i hope to speak/see you soon - all the best     marcus

ps please spread the word.

Kato Leonard schrieb am 25.11.21 um 15:36:

Hi @ZigBu https://github.com/ZigBu, thanks for explaining your process. Very intriguing, well done! As regards the license for ChemDraw, I installed it through the UCL software database https://swdb.ucl.ac.uk: search for ChemBio Office Ultra > Licenses > Personal Computer use information: click here http://informatics.perkinelmer.com/sitesubscription/. On the PerkinElmer site search for UCL and register for ChemDraw. In terms of the Maestro license, I don't remember how I got access to it since @Yuhang-CADD https://github.com/Yuhang-CADD helped me. Maybe he can explain it?

@ghostbio https://github.com/ghostbio same as Zige, located in London, GMT time. Many thanks!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/opensourceantibiotics/murligase/issues/58#issuecomment-979266009, or unsubscribe https://github.com/notifications/unsubscribe-auth/AMA3GKB5UQT5KAEXK26GCTTUNZCXHANCNFSM5IBNYMIQ. Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

--

Dr. Marcus Gastreich                @.*** Sr. Director Application Science Phone: +49-2241-2525-0 / Fax: -525               www.biosolveit.de BioSolveIT GmbH - An der Ziegelei 79 - 53757 St.Augustin - Germany facebook http://facebook.com/BioSolveIT                                            linkedin group http://www.biosolveit.de/linkedin

[drc007]

Might also be worth looking at https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8208308/

DeepFrag: a deep convolutional neural network for fragment-based lead optimization

https://durrantlab.pitt.edu/deepfrag/

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8243318/

[ZigBu]

Update on the allosteric binding site with 373 fragment derivatives created @KatoLeonard: All the compound is prepared by the ChemDraw then converted into the 3D version with minimized energy by openbabel. Then the docking process is done by the Autodock vina, all the results were analyzed visually. The configurations which resemble the originally 373 hit fragment and with the binding affinity less than -6.5kcal/mol were selected for further analysis. By aligning the crystal structure of the 3lk7, the Kato's compound and the 373 fragment can be compared visually in the PyMol. The screenshot of the results is revealed in the attachment. I use the command "align" and "pair_fit" try to align the compounds with 373, however both command fails and my estimation is that the size difference is too large for calculating the RMSD, thus I set 2 configurations as reference for the calculation of RMSD and the rationale is explained in the documents. Then the interaction between the compound and enzyme is analyzed by ligplot for reference. (DOI)

Docking.docx Ligplot interaction.pptx Please feel free to point out any errors in my results. Thank you so very much for any great suggestion Many thanks.Best wishes

[drc007]

@ZigBu Hi a quick question, do any of these compounds interact with Asp132? Or were the ligands not protonated on the piperidine nitrogen?

[ZigBu]

@drc007 Hi professor Chris Swain: Thank you so much for the question. During the docking process, I have added hydrogen to both ligand and 3LK7, since I am not an expert, I am not sure whether it indicates protonation or not. I would appreciate it if you could explain that. According to my observation, there is no interaction between Asp132 and the compounds. However, there are some interactions between Glu132 and amide nitrogen. Many thanks Wish you all the best!

[drc007]

@ZigBu I suspect if there no interaction with Asp132 then the ligands are unprotonated. It might be interesting to do the docking again with the ligands explicitly protonated.

[ZigBu]

Update on the allosteric binding site with 373 fragment derivatives created by @KatoLeonard: According to the suggestion from professor Chris Swain @drc007, all the ligands are protonated. Kato's molecule docking.docx Kato's molecule interaction.pptx

Based on compounds 1-11 created by Kato, these are the new compounds been created. According to the docking results, compounds 18-21 has the best performance and the synthetic processes are shown at the end of the attachment. Round2 molecule.docx

[mattodd]

OK, @ZigBu and @KatoLeonard - I'm conscious we need to capture these suggestions somewhere so that we don't forget them. How best to do that? Should we enter them as possible targets in the OSA Master List? Any reason not to do this, or does this represent something we need a compound registration system for @drc007? I should probably start a new Issue to discuss this again...

We also need to decide whether we are sufficiently happy with these molecules that we ask someone else to cross-check their docking with the target. Which should we go for? i.e. which subset are we happiest with, and which are reasonably simply synthetically accessible?

[drc007]

@mattodd @ZigBu @KatoLeonard I would not enter them on the OSA master list, we should keep that list for molecules that have actually been made. You could add another tab to the spreadsheet and call it "Ideas" and then add these to that tab. You could add additional information like who suggested it, a brief description why it was suggested and a link to the web page giving details. You could also include a note if someone decides to make the compound.

[danaklug]

We do have a Google from for submitting proposed molecules, although as it's set up right now you can only submit three at a time. Anyone can view the responses here.

It's very easy to modify the form so any suggestions about fields that should be included (or removed) are welcome!

[drc007]

@danaklug This looks great, Where are the results stored?

[danaklug]

@drc007 Right now they're stored on my Google drive which isn't the best solution. I can of course share the Google sheet with other Google accounts so we have talked about setting up an OSA Google account, but weren't sure whether that would be any easier in the long run. It would be great to get your input on that or if there's another way we can store things like this in the long term.

[ZigBu]

Conclusion：

In addition, I have found another study (DOI) where natural compounds were screened against drug target Acinetobacter baumannii MurD ligase. The in vitro experiment is conducted and the minimum inhibitory concentrations (MIC value) for the compound ZINC08879777 is between 100 and 200 ug/ml.