Docking - Science topic

Hello, thank you, we will take it into account in our future work.

Hello,

actually I have Autodock vina and I'm trying to use it but I can't find a tutorial/guide on how to dock copper and zinc ions to a protein. There is a tutorial called "docking with zinc metalloproteins", I think it works with only zinc metal ions though. Thanks for your response

Ankit Negi but it is possible if I use autodocktools or it can’t ?

I don't know which score you want to apply. However, docking scores usually do not take metal binding into account. For the reason, calculation such type of scores if you use is impossible.

Sargol Mazraedoost Thank you so much for your advice. I have solved the issues. :)

A sudden rise (very high value of RMSD) was followed by a sudden fall (very little amount of RMSD) in the deviation when plotted on the graph.

That depends on many factors like higher negative binding energies, binding pose and site, ADMET parameters of the compounds (Lipinski rule) etc. Based on these criteria you can choose your best compounds..

Regards

If your protein is a metalloenzyme with a metal in the active site, it would be expected to interact with a ligand that is a substrate or active-site inhibitor of the enzyme.

Jobin Thomas this is the pdbqt file of the docked protein-ligand complex.

Hello everyone. Did anyone get it solved? I am facing the same problem.

Ayaz Anwar Thank you for your contribution. Precise and great!

All noted.

Thanks.

Dear Ibrahim M. Ibrahim ,

You can use Zdock and ClusPro server as well to see if the results are useful.

Yes, and yes. You would have to provide which protonation state of your ligand from SMILES -> convert to 3D -> convert to pdbqt.

In Discovery Studio, after loading your docked protein-ligand complex, if it says no ligand molecules were found, you can solve this by following these steps:

By doing this, you should be able to view the interactions between your protein and ligand.

Dear all,

I am looking for a robust software to dock three protein molecules? I could not do it through Cluspro? Thank you.

: H

ow will you correlate between Mol Dock Score and Docking Score

One problem is that the pdb format is a very poor format choice for small molecule structural data. It is a very old format, dating back to times when computer memory was incredibly expensive. The .pdb format was developed to store crystallographic results for biological macromolecule structures, which use few standard building blocks, e.g. amino acids and nucleotides over and over again. It stores only the atom coordinates, and relies on the processing program to contain an internal dictionary providing the chemical details for each type of building block (residue), e.g. bond orders, charges, forcefield parameters etc., depending on the needs and purpose of that program. Display programs use distance criteria to draw the bonds between the atoms and get the bond order details from their internal dictionary. Docking, energy minimization and molecular dynamic programs for macromolecular structures similarely rely on internal dictionaries, but may allow you to extend these dictionaries as needed for modified residues and small molecules.

In contrast, it is simply not feasible to formulate an exhaustive dictionary for all possible small molecules. Small molecule chemical data file formats like .mol2 or .sdf need to store all information about the molecule (https://docs.chemaxon.com/display/docs/formats_mdl-molfiles-rgfiles-sdfiles-rxnfiles-rdfiles-formats.md ). The .cif format (crystallographic interchange format) is mainly designed to hold information about the crystal structure (such as unit cell values, atom names and their coordinates and any structural model quality indicators, e.g., R Factor) as well as any details of the diffraction experiment (such as temperature, pressure, experimental wavelength and the type and name of equipment used) and any data processing undertaken (such as the programs used to process the data) ( https://www.ccdc.cam.ac.uk/community/access-deposit-structures/deposit-a-structure/guide-to-cifs/#what-in-a-cif ). The related xpdb/mmCIF format is used by the protein database as a successor of the .pdb format

A .pdbqt file contains, in addition to the information from the pdb file, information on the charges (partial and full) on the different atoms. If this information can not be generated from the input and/or residue dictionary, no valid pdbqt file can be generated.

Yeah, in Pymol pretty easy:

Load the PDB file in Pymol:

> remove solvent (too many waters)

> remove polymer.protein (now you just have the ligand)

Maybe you just want the protein though

remove not polymer.protein (now you just have the protein)

Save the PDB files using the mouse.

The ligand has a 3 letter name in the sequence view, but usually if you just "select organic" then you have the ligand.

The difference in docking results between alpha-glucosidase and established inhibitors such as miglitol and acarbose is a complex issue that involves computational limitations, the intrinsic characteristics of the enzyme, and the physicochemical properties of the compounds. The unexpected superior performance of long-chain lipids highlights this issue. Molecular docking algorithms, such as GOLD and DOCK6, use scoring functions to predict the affinity between ligands and target proteins. These functions, although sophisticated, may not fully capture the complexities of molecular interactions, especially non-polar interactions and solvation effects. These effects are crucial for accurately modeling the behavior of lipid molecules. If interactions with long-chain lipids are not adequately accounted for, binding affinities may be overestimated.

Additionally, the active site of alpha-glucosidase is dynamic and can accommodate a wide range of substrates and inhibitors through substantial conformational adjustments, adding another layer of complexity. Docking algorithms that do not adequately model molecular flexibility may inaccurately predict the binding modes and affinities of relatively rigid molecules such as miglitol and acarbose, compared to more adaptable long-chain lipids. Furthermore, the hydrophobic nature of long-chain lipids drives them to interact with non-polar regions of proteins, reducing their exposure to water. If the docking algorithm gives too much weight to hydrophobic interactions, it may lead to incorrect predictions of higher affinities for certain molecules. This is especially true if the modeled alpha-glucosidase contains inaccurately depicted or overrepresented hydrophobic pockets.

To address these discrepancies, a comprehensive approach is necessary. This includes using enhanced sampling methods to better account for ligand and enzyme flexibility, utilizing improved scoring functions or empirical corrections that more accurately model hydrophobic interactions and solvation effects, and integrating experimental validation to corroborate computational predictions. Induced fit docking or ensemble docking techniques can provide a more realistic model of the ligand-enzyme interaction. Advanced scoring functions can offer a more nuanced understanding of the energetics involved. Ultimately, a synergy of computational refinement and empirical verification will be essential to resolve the apparent paradox in the alpha-glucosidase docking results and to advance our understanding of this enzymatic target.

Dear Professor Suvankar Banerjee

Thank you very much for your response and suggestion

Thank you!....rpalma@tectijuana.mx

It depends, if those het atoms are necessary for your proteins to interact, then nope, otherwise remove them. If these het atoms are part of crystallization process, you should remove them.

You need to prepare your system as well, make sure all atoms are present, geometry is optimized and protein is minimized. Pay attention to missing parts and secondary structure mismatch, just in case.

Once your proteins are ready, then go for docking.

Last point, avoid blind docking, search literature and see if binding residues are reported (if not for this protein, may be for homologous protein), and used them as starting point.

Assignment of force field parameters is essential step in any molecular modelling calculations. All the calculations are carried out using these parameters. From the error, it looks like that parameters/topology file for the indicated residues are in not available in CHARMM force field library. Given residue name are not canonical residue and hence their parameters have to be generated and appended as suggested by Dr Ayaz.

Sir, I added these parameters, but this line was already there in the dat file, Iam still getting the ERROR " Edit the parameter library file "AD4.1_bound_br.dat",

How different they are?

If you are focusing on a single ligand, it is not a good approach as you starting point is weak. Just look into the top 10/20 or 50 molecules from each docking program and compare them. How many of them are similar/same?

Just a suggestion, take a similar protein (same size or similar binding site) with known ligand (co-crystal structure) and do the docking exercise with them and see which program gives you better result. This can partially justify your results as well.

Yes, it is possible. That kind of simulation is called "implicit solvent molecular dynamics simulation". This simulation is notably much faster than the explicit one, albeit at the expense of reliability since the simulation doesn't reflect the real biological environment, as previously mentioned by Arnab Mukherjee .

To perform this, ordinary MD simulation software like GROMACS can be used and modified accordingly. Alternatively, commercial software such as MOE can also be employed for this purpose.

Add this line in the AD4_parameters.dat file

atom_par Se 4.21 0.291 14.000 -0.00110 0.0 0.0 0 -1 -1 1 # Non H-bonding

Regards

Waseem Ahmad Ansari I already checked the web tool you have mentioned none of the sites have the facility to dock ligands and receptors (docking of small molecules). Further, I also tried auto dock but no solutions.

hey Paul Shen

sorry to bother you but i have tried a lot and i didn't find the answer yet

I'm trying to use autodock GPU on colab but just for one ligand and and receptor

i find that my protein should be in .maps.fld format

but mine is in PDB or PDBqt could you help me if you find a way ?

There are some ways to dock with metal ions:

ideally we download the structure from https://www.rcsb.org/

Abdeen Tunde Ogunlana

Waseem Ahmad Ansari

Thanks for your response

Answering your initial question: no, it is not mandatory to optimize a molecule before docking. BUT you need to be sure that the structure you are using has no issue of clashing atoms, bad geometry, etc. Doing structure optimization with any molecular mechanics software (for example, OpenBabel) will be enough.

For your last question: if you are clear in the manuscript about using a good molecular structure for your ligands, there is no justification for asking to optimize it with any method (as it is expected that the docking protocol will change the initial conformation).

If you need more help, please, visit our Facebook group: https://www.facebook.com/groups/MolDocking

Did you install mgltools before installing autodock ? It contains certain commands that support autodock visualizations. You should try installing MGLtools and then re-install autodock. with updated versions

As previously suggested, the PDB is a textual file. You can open it and make all the modifications you need. Be careful not to misalign the columns to avoid disrupting subsequent calculations.

If you have Gromacs installed, you can renumber the PDB file using the command: gmx editconf -f file.pdb -resnr starting_number -o output.pdb

Replace file.pdb with the name of your PDB file, starting_number with the desired starting number for residue numbering, and output.pdb with the desired output file name.

Prasenjit Bhowmik Thank you soo much. I'll look into both the softwares.

hello Yashee Gupta ,

grid> output> save gpf.

this save file in file format so autogrid can not browse that file.

for get gpf file go on that file > right click> properties> add .gpf extension on their name so that file will be convert in to .gpf file.

Bartosz Trzaskowski Thank you. Could you please explain how I could modify the Autodock forcefield to include parameters for this element?

By the above code, irrespective of protein size the grid box size will be considered as 20x20x20. End of the vina execution, most of the complex shows binding affinity "0" or much less, as the active site will be out of the grid box range. Better increase the grid box size (SIZE_X,Y,Z) up to 60 or 120 each, depending on the maximum proteins (chains in PDB code) size of each complex, and try to run VINA again. Then you may get binding energy values of maximum protein-ligand complexes (sometimes for all).

However, this will not mimic the experimental structure correctly, since you handling bulk protein-ligand (separately) complexes docking with the common configuration file same time.

H K Ra Please read Gibbs Free energy and its relation to protein folding. You will for sure taste the whole food.

Few ligand docking software allows you to explore multiple poses for a given ligand in a specific binding site. While CB-Dock2 might not provide this feature directly, you can try other docking tools like AutoDock or GOLD, which allow for flexible ligand docking and can provide multiple poses.

Here are some popular tools you can consider:

It is important to carefully analyze the results from these tools and consider using multiple tools to cross-validate the predicted binding sites.

I think with the help of LigPlot+, you can see the 2D protein-protein interaction.

But in any case, you have to change the chain identifier of one of the proteins. you can manually edit the chain identifier directly in the PDB file using a text editor. Locate the lines that define the chain identifier and change the chain identifier accordingly. Save the modified PDB file after making the changes.

You should choose the pose with the highest score (-10.1) since it has a significant score difference with the second pose.

Docking algorithms like Autodock Vina evaluate multiple factors when calculating the docking score, including van der Waals interactions, electrostatic interactions, and hydrogen bonding. While the number of hydrogen bonds can be an important indicator of ligand binding, it's not the sole determinant of the docking score. Other factors such as the strength and geometry of the hydrogen bonds, as well as non-polar interactions, can also contribute to the overall score.

Alisa Gorislav Thank you !

Dear Researcher, Cancer is a polygenic disease, numbers of biological target (proteins) associated with skin cancer or melanoma....Some are BRAF, MEK1 & MEK2,PD1 etc ..Please firstly confirm your aim and objectives of research, do some literature survey to convey biological target responsible for skin cancer then retrieve it's structure directly from PDB and check ligand receptor interaction...

Hello Punet Kumar, As Satyendra Singh highlighted, PyMol primarily functions as a visualization tool and lacks inherent docking capabilities unless supplemented with extensions. If your goal is to conduct docking experiments involving multiple ligands with a single protein in a single run, I recommend considering AutodockVina. This software is both free and open-source, providing a robust solution for your docking needs.

Thank you very much for your reply.

Well, my astute researcher friend Geetmani Singh Nongthombam, let me break it down for you Geetmani Singh Nongthombam. In the realm of ligand docking, the accuracy of binding at a specific protein site is crucial. While it's not mandatory for the grid box to be localized precisely where the reported inhibitor binds, it certainly helps enhance the reliability of predictions.

Now, about blind docking versus knowing the binding site from XRD structures, it's a bit like navigating uncharted waters versus having a detailed map. Blind docking explores the entire protein surface for potential binding sites, which can be a shot in the dark but might unveil hidden gems. On the other hand, if you Geetmani Singh Nongthombam already know the binding site from experimental XRD data, you're like a captain steering straight for the treasure. It's more targeted but doesn't account for potential surprises elsewhere.

In the world of molecular interactions, as in life, sometimes it's about calculated risks and exploration. What approach you Geetmani Singh Nongthombam choose depends on your appetite for adventure and the specific goals of your research voyage. Smooth sailing!

Map file create after autogrid. Maybe problem in grid preparation thus, map file not generate well

Thank you for your response. Is it possible to dock one ligand, and then, taking into account the influence of the first ligand, proceed to dock the second ligand (same ligand) in the presence of the first one? @Sargol Mazraedoost

I will agree with the first explanation of @Hamed Shahriarpour. However, his second explanation looks invalid. While we can analyze multiple ligands, there is no competing interaction between the ligands. By the time the analysis of the first ligand ends, the binding side returns to default with no information about the first ligand.

Pyrx is a great program

Marjanur Rahman Bhuiyan Thank you for the answer.

I get stuck in analysis part, like when we do protein ligand docking, in that case we can analyse and visualize the interactions using pymol or discovery tools but how are we gonna do this in case of protein protein dockiing?

Another question is that in the case of ligand and protein, ligand sits very efficiently in the active site of protein as the ligand structure is small, but in case of protein protein, both structures are very large and what i think is that in that case protein will not be able to sit in the active site so in that case should we just consider the interaction of protein with the active site of another protein?

Hi,

I think it depends of several parameters. Docking score function only represents how "stable" or how much favored your proposed binding mode is, in terms of energy. Although this could serve as a good approximation, or even as a first screening method, the protein is generally considered as rigid, which does not seem so much reliable.

In the other hand, RMSD brings information about how flexible is the ligand once it is bonded to the target. Low values mean a stable behavior within the pocket, but personally I should consider the option of carrying out a MD simulation just to check which way the system undergoes and after that, evaluate the RMSD values during a trajectory.

Remember that docking poses should be just a starting point for subsequent computational analysis, so a proper selection of that must be a priority. Check the main interactions showed by each pose, compare them among all the ligands you have and then, try to perform more computational procedures to hold a more accurate analysis.

Bests,

Dear Vishal Zambre,

Make sure about your PDB and ligand preparation before Docking. Issues in ligand preparation, such as incorrect atom typing, atom order, or charges can lead to flipped poses. Also check symmetry and chirality, because in some cases, docking algorithms may produce inverted chiral centers. Ensure correct stereochemistry in both co-crystallized and docked poses.

Adjust flexibility, scoring, and search parameters for accurate poses and also avoid over-optimization to prevent false positives. If the binding pocket undergoes substantial conformational changes upon ligand binding, prefer flexible docking or induced fit docking approaches.

You can verify input coordinates and formats for both co-crystallized and docked structures. Additionally, try alternative docking software to assess if the issue persists, as different algorithms may yield varied results.

Conduct post-docking minimization or molecular dynamics simulation to refine docked poses and address inaccuracies introduced in the docking process.

Carefully examine these aspects to identify the cause of flipped poses and implement necessary corrective measures.

Thank You

Did you find the solution to solve this problem? I have the same error and try to use another Windows system, but it's useless.

Hello. In order to dock a protein whose 3D structure is not available on the uniprot site, you must model its 3D structure, for this there are many servers such as Itasser, Quark, and Robetta, and after evaluating and measuring your 3D structure, you can do the docking by Servers like Cluspro, Zdock or Hdock

it's important to consider specific proteins that play crucial roles in the survival and reproduction of microorganisms. Enzymes involved in cell wall synthesis: Proteins like penicillin-binding proteins (PBPs) are crucial for bacterial cell wall formation.

DNA gyrase and topoisomerases: Involved in DNA replication and repair, these are essential targets for antimicrobial compounds.

Ribosomal proteins: Targeting bacterial ribosomes can disrupt protein synthesis. Utilize databases like the Protein Data Bank (PDB) to find crystal structures of your selected proteins. Molecular docking predictions should be validated through in vitro and in vivo experiments.for in vitro evaluation you can use microorganisms directly.

Some factors to consider when comparing results from different methods:

1. Methodological differences: Docking and free energy calculations using ONIOM are based on different underlying principles and assumptions. Docking typically predicts the binding pose and affinity of a ligand to a receptor based on scoring functions and geometric complementarity. ONIOM calculations, on the other hand, use a combination of quantum mechanical (QM) and molecular mechanics (MM) methods to describe the electronic structure and energetics of a system. These differences in approach can lead to variations in the absolute values obtained.

2. Relative comparisons: As you correctly mentioned, the primary utility of comparing results from different methods lies in their relative comparisons. The goal is to assess the relative binding affinities or trends between different ligands or receptor-ligand complexes within the same method. For example, if docking experiments consistently rank a set of ligands in a certain order of affinity, and the ONIOM calculations show a similar trend, it provides confidence in the relative comparison between ligands using both methods.

3. Calibration and validation: To establish a meaningful correlation between different methods, it is often necessary to calibrate and validate the results against experimental data or reference datasets. This can involve benchmarking a set of compounds with known binding affinities and comparing the relative rankings obtained from different methods. Calibration can help adjust the scaling factors or parameters to improve the agreement between methods and experimental results.

4. Limitations and assumptions: It's important to consider the limitations and assumptions of each method. Docking, for instance, may not fully capture the precise energetics of the binding process, while ONIOM calculations may have limitations in terms of the chosen level of theory, basis set, or representation of the system. Understanding these limitations can guide the interpretation and comparison of results.

Hope it helps:credit AI

I'm also having the same problem with my predicted model.

Here are a few popular docking tools that allow for multiple-ligand docking:

When dealing with protein-protein docking involving non-standard amino acids, it's important to choose a computational tool that supports these non-standard residues. One such tool that can handle non-standard amino acids is HADDOCK (High Ambiguity-Driven Biomolecular DOCKing).

HADDOCK allows for flexibility in defining molecular interactions, which can be crucial when dealing with non-standard amino acids or modified residues. It supports the use of user-defined restraints, making it adaptable to various types of molecules and molecular features.

Here are the general steps you might follow with HADDOCK:

Prepare Input Files: Prepare PDB files for both interacting proteins, including the non-standard amino acids. Define any additional parameters required for the non-standard amino acids. Define Active and Passive Residues: Specify the residues that are considered active or passive in the binding site. Run HADDOCK: Submit the job to the HADDOCK server or run it locally. Analyze Results: Analyze the docking results, considering the interactions involving non-standard amino acids.

When this type of error occurs to me, I prepare the ligand file again using the openbabel software, without making any modifications... just rewriting it.

example:

obabel -ipdbqt lig.pdbqt -opdbqt -Olig.pdbqt

Typically, such a solution works.

To analyze and validate docking procedure for a protein with no co-crystallized ligand, you need to find the key residue in the active site or binding site of the protein. After choosing the best pose based on docking score and interactions with key residue, you can perform molecular dynamics to validate the docking procedure.

The reasons for control docking/ redocking more than 2 Å might due to:

1. Not included all the important amino acids in the box.

2. The protein protonation at the pocket or ligand charge, but if you follow the standard protocol, it should be no problem.

3. You might need to check for ligand interactions. If there is a covalent bond, you might need to use the covalent bond protocol.

To get the RMSD below 2, you might need to reevaluate the box size/parameters.

Vikas Kumar bro how did you solve this problem? could you please give me your email in the inbox so that I can communicate with you

Thank you Waseem Ahmad Ansari and Soykan Agar for your answers. But I could not solve the problem. Can you kindly suggest how to choose the ligand as a static molecule?

Hello Mr. Thomas

Thanks for your valuable comment.

Hello

Here are some common issues and suggestions to troubleshoot:

Hi Kwasi Kantanka Safo ,

This question has been answered before. You can check https://www.researchgate.net/post/How-can-I-parametrize-an-atom-for-Autodock-Tools

If you mean this ClusPro https://cluspro.org/ that you want to use?

It is said clearly that it is a protein-protein docking server. Metformin is not part of protein structure, it is a small molecule ligand.

Arkajyoti Banerjee

My version of YASARA (YASARA-Structure) requires a paid license. However, the price is very reasonable compared to other comparable software such as Schrodinger or CCG MOE. In addition, the licensing policies are extremely liberal. For example, I can run as many simultaneous instances as I wish, on the same computer and/or on multiple computers with the same or different operating systems (it runs on Linux, Mac, Windows, and Android).

There is also a free "viewer" version of YASARA. I have not tried the free version, and I do not know its capabiliites, but I suspect that it would not include docking.

I am glad to know that you got good results with classical AutoDock 4. Sometimes, it works out that AutoDock 4 gives better results than Vina. It depends in part on the particular receptor-ligand system. Standard AutoDock 4 is slower than Vina, although now there is an accelerated version, AutoDock-GPU, than runs on OpenCL or CUDA to achieve very fast speeds.

Hello Muritala Olusola

In the event of the appearance of this issue at the existence of all required maps in your directory, you can also manually check the content of your .dpf file and verify the names of all maps.

I had a similar error with the desolvation map. I used the long name for my .d.map file, and when I opened the .dpf file I found that it contained the following description of the desolvation map: "desolvmap 1qkm_wo_non-std_rsdus.d.map# desolvation map". So, the name of my .d.map file took the whole space up to the "#" sign, and probably AutoDock4 wasn't able to recognize the correct name of the file. The addition of a single space between the name of the map file and the "#" sign as follows "desolvmap 1qkm_wo_non-std_rsdus.d.map # desolvation map" resolved my problem.

Hi Fatemeh

maybe you can try HDOCK or PATCHDOCK webservers.

Here are some factors that can contribute to differences in docking results:

Software and Versions: Different docking software packages may use distinct algorithms and scoring functions. Even updates to the same software can lead to differences in results.

Scoring Functions: The scoring functions used to evaluate the binding affinity of ligands to the protein can vary between software and versions. Different scoring functions may prioritize different types of interactions (e.g., hydrogen bonds, hydrophobic interactions) or use different energy terms.

Ligand and Protein Preparation: The way ligands and proteins are prepared for docking, including the addition of hydrogen atoms, assignment of charges, and optimization of structures, can affect the results.

Grid Settings: The definition of the docking grid or search space can have an impact. Slight variations in grid dimensions and placement can yield different results.

Conformational Flexibility: The treatment of ligand and protein flexibility can also influence results. Some docking programs allow flexible ligands or proteins, while others keep them rigid.

Optimization: As you mentioned, optimization of ligands can significantly affect results. Different optimization methods and parameters can lead to variations.

Parameter Settings: Docking programs have numerous parameters and settings that can be adjusted. Small changes in these parameters can lead to differences in results.

Initial Position: The initial position of the ligand within the binding site can also influence the outcome. Docking software often performs multiple runs with different initial positions.