Difference observed in protein RMSD between different complexes but protein is same

User discussions
,
1

GROMACS version: 2023.3
GROMACS modification: Yes/No
Here post your question

Am completed MD Simulations of a single protein with different drugs and 8 complexes are formed with same protein. But in RMSD analysis why my protein RMSD is difference in all complexes and protein is same. How am represents the graph for a protein with all complexes RMSD in single figure. which complex of protein is i need to select for comparision with different RMSD complexes ?

please give me the suggestions.

image
image845Ă—578 28.6 KB

2

To clarify, are you asking why the RMSD graphs for the different complexes are different despite being the same protein? If so, the graphs aren’t going to be exactly the same since every simulation involves some randomness when generating the trajectories of each atom.

Or are you asking why the RMSD is non-zero even though the complexes didn’t change from the original protein? If so, the RMSD is comparing the atoms of the complex that are moving in the simulation to whichever static structure you chose, so there’s going to be some amount of RMSD no matter what. As long as the RMSD flattens at a small value and you don’t see anything change in the actual simulation, that suggests stability.

Which protein to compare the complexes to depends on what you’re trying to find out. If you want to see how different each complex is to the protein without the drug, then you compare it to the protein without the drug. If you want to see how stable the complex with the drug is overtime, you compare it to the starting structure of the complex.

Hope this helps, feel free to ask a follow-up if I didn’t understand the question correctly!

3

Thank you Karis for your response,

Actually am running MD Simulations for one protein with different drugs. After I have 8 complexes in different drug with same protein but Analysis of RMSD I found a doubt that is if am compare 8 complexes RMSD’s with protein RMSD which protein of RMSD is am taken for comparison. The below graph shows same protein RMSD in different complexes.

  1. Why the protein RMSD is different in all complexes (protein is same in all complexes)?

  2. Which protein RMSD am taken for comparison of complexes (pro-lig)?

4

  1. Not sure I understand, you’re asking why Apt_pro, Lor_pro, RAF_pro, PON_pro and NTD_pro don’t all have exactly the same RMSD in the graph even though they are the same protein just with different drugs, is that correct? Every simulation generates the starting velocities of the atoms randomly so even the same protein won’t have the exact same rmsd for different simulations, as long as the overall trend is the same it shouldn’t be a problem

  2. Also not sure I understand, are you asking which protein you should compare the complexes to for the RMSD analysis? As mentioned above, that depends on what you’re trying to find out.

Sorry if that’s not answering the question, I might need you to rephrase it a bit more

5

Thankyou Karis

Your answer is helpful for my work Karis. You said right, all protein RMSD of complexes are not getting the same peaks. In the figure below, Is one protein RMSD for my research not a problem for other complexes?

6

Not having the same peaks doesn’t matter as long as the overall trend is stable.

Sorry in advance, but for that last question, could you rephrase? I’m not sure what you mean by one protein RMSD not being a problem for the other complexes.

7

The image in 5 proteins (same protein) of different complexes are there. if i choose any one of RMSD complexes comparison is not a problem?

Like

  1. protein_1 vs Drug_1
  2. protein_1 vs Drug_2
  3. protein_1 vs Drug_3
  4. protein_1 vs Drug_4
  5. protein_1 vs Drug_5

These are my complexes same protein in all complexes. The image in proteins are belongs these complexes so if i choose anyone RMSD of protein for comparison is right or wrong?

Thank you for your response Karis

8

I’m not sure what you mean by choosing one of the RMSD complexes for comparison, what are you trying to compare the RMSD to? Is there a reason you need to choose only one of them?

9

Am trying for comparison between protein RMSD (Apoprotein) with complex RMSD’s . but my problem is every complex in protein RMSD is changed and that complexes in only drug is different but protein is same for all complexes. Now am trying to compare protein RMSD (Apoprotein) with complex RMSD’s . In that case i identify every complex in protein RMSD is changed is shown in the figure. that 5 RMSD’s are related only one protein (not different one) and it came from different complexes. So my question is

  1. If i compared RMSD of protein and complexes which protein (image in protein from different complexes) is choose from comparison?

  2. this type of comparison is right or wrong ?

10

To clarify, you want to compare the RMSD of the complexes (protein + drug) to the protein without the drug, and you’re asking if it’s alright to choose one of the complexes to represent the protein on its own and compare the other complexes to it, is that correct? If the goal is to compare the complex to a protein without the drug, I would think running a separate simulation of the protein without the drug and getting its RMSD would be the best way to do it, unless I’m still misunderstanding the question.

It might help if you explained what exactly you’re trying to find out from the data, it’s still a little unclear to me what you’re trying to compare and what question you’re trying to answer

11

Karis, thank you for your patience to understand my problem.

My protein is hemoglobin and my drugs are Apa, Lor, RAF, Pon and Nin.

Now I am running 1. MD simulations for hemoglobin (protein) with Apa(drug)

  1. MD simulations for hemoglobin (protein) with Lor(drug)

  2. MD simulations for hemoglobin (protein) with RAF(drug)

  3. MD simulations for hemoglobin (protein) with Pon(drug)

  4. MD simulations for hemoglobin (protein) with Nin(drug)

After MD simulations I found 5 complexes

  1. Heamoglobin-Apa

  2. Heamoglobin-Lor

  3. Heamoglobin-RAF

  4. Heamoglobin-Pon

  5. Heamoglobin-Nin

Now I would like to compare hemoglobin (protein without ligand) RMSD with the 5 above complexes RMSD.

But my question is a running simulation for each complex of protein (hemoglobin). Heamoglobin-Apa complex in took protein(hemoglobin) rmsd, heamoglobin-Lor complex in took protein(hemoglobin) rmsd, heamoglobin-RAF complex in took protein(hemoglobin) rmsd, heamoglobin-Pon complex in took protein(hemoglobin) rmsd, heamoglobin-NIN complex in took protein(hemoglobin) rmsd.

In this case protein hemoglobin is common in all complexes only the drug is different but why am I getting a different RMSD plot of protein (hemoglobin) from each complex but protein is same and common in all complexes. The image is also said to be the RMSD of protein (hemoglobin) from each complex after running MD simulation for the lone protein.

  1. Why is the protein RMSD changed in every complex but the protein is the same?

  2. Now which one of hemoglobin protein (only protein) RMSD is chosen for comparison with all complexes?

Note- The graph is shown 5 plotes are apoprotein(hemoglobin) RMSD from 5 different complexes and I run sepearatly simulation for each (5 ) complex to get protein alone (hemoglobin) RMSD and the plots are not a complexes RMSD and they are protein RMSDs of hemoglobin from each complex.

12

image
image958Ă—667 36 KB

13

For the first question, as mentioned before, multiple RMSD plots of the same protein not being exactly the same is to be expected, the velocities for every md simulation involves some randomness in how the initial velocities of the atoms are generated.

When you say “The image is also said to be the RMSD of protein (hemoglobin) from each complex after running MD simulation for the lone protein”, is this saying you first ran the simulation of the protein in complex with the drugs, then removed the drugs and ran the simulation for those proteins on their own to get the RMSD graphs? And is the second question asking which one to use to compare to when the drugs were still in complex? In that case, I’m not sure it makes sense to pick one of the RMSD plots to compare.

  • If you want to compare the complexes to just the protein that hasn’t been in a complex to see what effect the drug has while it is bound, you would run a separate simulation of the protein without ever having a drug in complex and use that to compare the RMSD.
  • If you want to compare the complexes to the protein once the drug is removed to see what effect the drug might have on the protein after it unbinds, you would compare the RMSD of the protein in complex only to the RMSD of the protein that used to be in complex to the same drug

Unless there’s a reason you’re picking out one complex in particular, I’m not sure it makes sense to run a simulation for all 5 cases where the drug is removed from the complex only to choose 1 as a point of comparison.

A more general point I want to make, keep in mind that RMSD doesn’t tell the whole story, it just compares the difference of the molecule throughout the simulation to whichever static structure you chose, there’s some significant difference between the structures, you might not notice it just from RMSD

As always, feel free to follow-up if I’m still misunderstanding the question

14

Thankyou so much Karis,

I am learning good knowledge from you.
I am not running MD Simulation for protein alone and after running simulation for protein_ligand than am taken protein and complex rmsd from below the options

(base) venukrishna@DESKTOP-PIBELR0:/mnt/c/1vr2_6gqo_Vs_eight_drugs/6GQO_RAF_MD_100ns$ gmx_latest rms -s em.tpr -f md_0_10_center.xtc -n index.ndx -tu ns -o rmsd_protein.xvg :-) GROMACS - gmx rms, 2023.3 (-: Executable: /home/venukrishna/gromacs-2023.3/build/bin/gmx_latest Data prefix: /home/venukrishna/gromacs-2023.3 (source tree) Working dir: /mnt/c/1vr2_6gqo_Vs_eight_drugs/6GQO_RAF_MD_100ns Command line: gmx_latest rms -s em.tpr -f md_0_10_center.xtc -n index.ndx -tu ns -o rmsd_protein.xvg Reading file em.tpr, VERSION 2023.3 (single precision) Reading file em.tpr, VERSION 2023.3 (single precision) Select group for least squares fit Group 0 ( System) has 55392 elements Group 1 ( Protein) has 4956 elements Group 2 ( Protein-H) has 2475 elements Group 3 ( C-alpha) has 308 elements Group 4 ( Backbone) has 924 elements Group 5 ( MainChain) has 1231 elements Group 6 ( MainChain+Cb) has 1519 elements Group 7 ( MainChain+H) has 1524 elements Group 8 ( SideChain) has 3432 elements Group 9 ( SideChain-H) has 1244 elements Group 10 ( Prot-Masses) has 4956 elements Group 11 ( non-Protein) has 50436 elements Group 12 ( Other) has 53 elements Group 13 ( LIG) has 53 elements Group 14 ( CL) has 1 elements Group 15 ( Water) has 50382 elements Group 16 ( SOL) has 50382 elements Group 17 ( non-Water) has 5010 elements Group 18 ( Ion) has 1 elements Group 19 ( Water_and_ions) has 50383 elements Group 20 ( Protein_LIG) has 5009 elements Select a group:

  1. It is right or wrong for comparing RMSD of protein and complex from above options? and taken both protein and complex RMSD’s from above option after completion of protein_ligand MD simulations.

  2. I am not running MD simulation for protein alone and could you please tell how to run MD simulation for only protein without ligand?

15

Screenshot 2024-02-28 173109
Screenshot 2024-02-28 1731091355Ă—555 17.2 KB

16

  1. So for each complex, you want to get the RMSD graphs from the same simulation, 1 graph for only the protein (so choosing group 1), 1 graph from the protein + ligand (so choosing group 20), and compare the RMSDs, is that correct? You could do so, but I’m not sure what conclusions you could draw from it. Both groups would compare the same protein to its starting structure from em.tpr, so the only difference would be from the ligand RMSD, which you could just get separately by choosing group 13. The RMSD between em.tpr and md_0_10_center.xtc only tells you how different the protein or complex is from the structure you began with, so I’m not sure what comparing different parts of it will really tell you.

  2. I don’t know how you prepared the system to begin with, so:

  • if you started with the proteins and ligands separated and then put them together yourself to make the complexes, then of course you could just run with the protein file.
  • If the file you started with comes with the protein and ligand already in complex, but there aren’t any bonds between them, then you could extract a .gro file using gmx editconf -f input.gro -n index.ndx -o output.gro, and select the group containing the protein, then use that to run the simulation.
  • If the file containing the complex is made in such a way that there are bonds, that may throw up an error when you try it this way since the bonds mean the structure of the protein is different with and without the ligand, in which case you might need to manually fix the protein by fixing the parts which were bonded to the ligand. (If this were the case, you would have had to generate the parameters of the protein bonded to the ligand yourself to have gotten this far, so you probably don’t have to worry about this option)
17

Thank you Karis,

  1. For example If I compare RMSD’s of an apoprotein and complex then separately I need to run MD Simulation for both protein and complex?

or else

  1. then choose option 1 for protein RMSD and 20 for complex RMSD. Is it ok for RMSD’s comparison after performing MD simulation of protein_ligand complex?
18

To be clear, RMSD just measures how closely the protein or complex resembles a chosen structure. In this case, the RMSD is measuring how close the molecule during the simulation is to the structure you got from the em step. The RMSD of the protein is measuring how closely the protein during md_0_10_center.xtc resembles the protein in em.tpr and the RMSD for the complex is measuring how closely the complex during md_0_10_center.xtc resembles the complex in em.tpr. A low RMSD just tells you the structure is staying close to what it started as.

What I’m saying is, how you compare them depends on what you want to investigate.

  • If you want to know the effect which the drug has on the stability of the molecule (maybe you want to see if the protein is stable without the drug then changes shape with the drug, or the other way around), you would want to compare the protein with and without the drug, so you would want 2 separate simulations, one for with and one for without the drug, because If it were the same simulation, the effect of the drug on the protein is still there even if you choose group 1, it’s just that the atoms in the drug itself is not included when calculating RMSD.
  • If you want to know how being in a complex has a different effect on the protein than it does on the drug (maybe you want to see if the protein is stable and the drug changes shape when they bind, or the other way around), then you could use the same simulation of the complex to get the RMSD of group 1 and group 13 to compare them

I’m just not really sure what you would get from comparing the RMSD of group 1 and group 20 of the same simulation

19

Now am understood karis thank you for your explanation

I want to know which drug is stable in protein and If I compared five complexes with protein. Could you please tell me about how to run MD Simulation for protein? a brief explanation about MD simulation of protein. Can you provide any links and literature for that simulation.

20

It should be the same method you used to run the simulation for the complexes, just with a starting pdb file that only contains the protein rather than the whole complex. If you need a general guide, these tutorials should be helpful, particularly the first one for the basics on running simulations. If you need to do something specific that isn’t described in the tutorials, you could refer to the documentation