Hello. I ran MD for a protein complex, consisting of a ligand protein with 380 amino acids and an extracellular domain of a receptor, consisting of a protein with 192 amino acids. I`ve completed the MD process based on the last article of GROMACS. before analyzing, I did some trjconv processes and after that, RMSD and RMSF were calculated. they are not good diagrams, I really dont know how to fix it. I sent the photos of the diagram and the trjconv cods. can you help me? 1- gmx trjconv -s prep/md.tpr -f prep/md.xtc -o analysis2/md_nojump.xtc -pbc nojump system
2- gmx trjconv -s prep/md.tpr -f analysis2/md_nojump.xtc -o analysis2/md_centered.xtc -pbc mol -ur compact -center protein system
3- gmx trjconv -s prep/md.tpr -f analysis2/md_centered.xtc -o analysis2/md_fit.xtc -fit rot+trans backbone protein
4- gmx rms -s prep/md.tpr -f analysis2/md_fit.xtc -o analysis2/rmsd.xvg -tu ns backbone backbone
5- gmx rmsf -s prep/md.tpr -f analysis2/md_fit.xtc -res -o analysis2/rmsf.xvg c alpha
How are your residues numbered in both proteins? What about chains, are the proteins considered as one chain or as two separate?
two chains, each numbered from 1. you say that are these bad diagrams because of this?
Yes, in the case of RMSF, XmGrace is trying to plot two different data series as one continuous series. You’re just getting a bizarre connecting line there, but the actual data appear normal.
is it really normal? i thought the fluctuation is not normal somewhere, especially because it is a protein and ligand in pdb and I expected better stability
There is no way for any of us to comment on whether the results make sense or not, but there is nothing in either plot that looks unusual tom me other than the plotting glitch. Regions with large RMSD and/or RMSF are often flexible loops, so you need to try to interpret your results in light of what you know about the protein structure and what you observe from the trajectory. The data shown do not include any aspect of ligand dynamics, just protein structure. Note, too, that crystal structures are not necessarily “ground truth” when the protein is placed in solvent, and loops and other motifs may change quite dramatically.
That level of RMSD can be actually pretty normal for a structured protein that may have many flexible loops, as said above. I previously wrote that this deviation could be even low, but I would like to clarify that only in the very flexible regions! sorry about the lack of clarification. Now, if the protein is very globular and compact that would be a different story.