Hello. In the process of energy minimization, there is a hint as shown in the figure. And I found that the results were really very incorrect. I wonder how I should find the overlapping atoms (there are 570) and modify them.
570 overlapping atoms? How did you generate the system? If there’s only one or two atoms you can usually modify the coordinates by hand. But with that many I’d have a look at the setup process to see how it happened.
I have solved the problem, thank you. There was an error in my PDB file. It has now been corrected. Thanks again.
What was the error in your PDB file if you do not mind sharing as I am having the same problem.
Hi Magnus,
I am encountering the same problem but with 5 pairs of two atoms. I found this number by creating a python script that read my box.gro file and outputted which set of atoms where within 0.1 units of each other in the x, y, and z coordinate. I created my peptide using Chimera and then generated my system with AMBER99 force field. Where do you think the problem is occurring and how do I manually modify the coordinates by hand while maintaining the stability of the peptide.
It is difficult to say where the problem occurs, but I would suspect that Chimera might not generate that good coordinates for peptides, especially not if they are complex or large. Another alternative would be that there is overlap across the periodic boundary, i.e., that the box is slightly too small for the system. But that should be easy to find out.
I have tested the periodic boundary condition possibility by running my energy minimization with an increased box size up to 5Å. While some sequences of my peptide are running successfully, others are not, which is making it challenging to pinpoint where the problem is occurring. It’s important to note that I am working with an ELP (Elastin-Like Polypeptide) beta-spiral model, specifically the URRY sequence, which is known for forming beta structures.
In my workflow, I’ve been adjusting the torsion angles of my peptide using Chimera to explore different conformations. This approach is not uncommon when working with beta-spiral models, as these torsion angle adjustments can help mimic the natural flexibility and variability observed in such structures. However, given that some sequences are not minimizing successfully, I am concerned that there may be an issue with the torsion angle adjustments or another factor that I haven’t yet identified. Given this context, I would appreciate any insights or suggestions you might have on resolving this issue. Could there be specific considerations or common pitfalls related to the URRY ELP beta-spiral model that I should be aware of when modifying torsion angles?
5Å sounds like a very small box size. When generating the system, generate a box around the peptide, e.g., gmx editconf -d 0.9 -bt dodecahedron. After that, solvate the system with gmx solvate. Avoid modifying the box size after solvating the system - unless you really need to.
Or is the polypeptide periodic, i.e., repeating itself across the periodic boundary? In that case you can’t use the -d option, but you must set the box size corresponding to what would be suitable for your molecule. 5Å sounds very small for most force fields. I haven’t worked with URRY ELP beta-spiral model, so I’m afraid I can’t give any specific advice.