amino acids identity is 95%, 66.09% of structural similarity (known 434 residues and unknown 344 residues). One already predicted known protein is using for industry purpose, so can I suggest another one also can use for industry purpose?
A C-alpha backbone RMSD of 0.302 Angstrom indicates that the structures are essentially identical by this criterion.
Hi Karthika,
With structural and sequence similarities you have to be very caution. Imagine you have just one mutation it the active site, and your protein would be not active. You have to check carefully an active site, and make sure that all important residues are in proper positions. Your overall RMSD could be low, but active site residues could have very different layout. The same with the sequence identity, few residues in the important locations may turn your Enzyme into non-active protein.
If the structure you are comparing with has any publication with the mechanism of the Enzyme, I would recommend to read it carefully and check active site of your protein.
The best would be to run assay, if it is an easy assay.
Good luck!!!
George.
Hi Karthika,,
from a structural point of view, a RMSD of 0.302 A don't meant perfect superimposed, just quite close folding. For instance, at this level of RMSD you could expect to be lucky and have a solution in molecular replacement method. Also you must take in account the final number of C-alpha superimposed between both structures. Best
Structural similarity.
"Perfectly superimposed" (the same structure) would give an RMSD of 0.0. Of course in the real world there are errors & uncertainties in the data, so RMSD = 0 would not be achievable if for example you compared a X-ray structure of the same protein done on a Tue or Wed (you would have to use 2 different crystals because zapping them with an X-ray beam kills the crystal). So an RMSD of 0.2 to 0.3 is very close, about what you would get for the example of comparing 2 crystals of the same protein above. (The error in individual dihedral angles for any X-ray or NMR structure is at least 10-20 degrees.) The RMSD also depends on the number of atoms compared - a low RMSD for 7 residues is not that significant, because an alpha-helix in lysozyme will look like an alpha-helix in hemoglobin (RMSD improves as the size of what you compare decreases - the russian doll effect)
Hi,
There are two questions: one in the title and one in the text. First question is a technical. We have to remember one fact, that in crystallography we are dealing with space in grids. Our real space in crystallography is NOT continues. Most structure refinement programs use grids at 1/3 of data resolution. Which meant, that for the structure refined to 2.0A resolution, spacing between grids will be about 0.7A. This will let us resolve two spots at 0.35A. If your structure is refined to 1.5A, you will be able resolve spots at 0.25A distance. I am not taking in consideration mosaicity or optical resolution of the data. So, technically RMSD of 0.3A between two compared crystal structures at 1.5 - 2.0A resolution are within uncertainties in the refined positions of atoms of each structure. So, the answer for the first question is YES, two structures are identical.
But it does not mean that proteins will perform the same reaction the same way. As I mentioned earlier, just one point mutation in the active site and your enzyme is dead. Bacterial resistance is due to just few mutations in the active site of some specific proteins. Some mutations in the active site may improve the reaction, others will just shut it down.
Which makes scientist involved in the structural analysis to be very careful about conclusions. Just one number, in this case RMSD, will not give one a compete picture. One have to carefully examine the "small" details, and make sure that "a tail they pulled out is not attached to an elephant".
Best wishes.
Thank you very much for your all valuable answers. I learned a lot
As an alternative to RMSD, you could also consider structural superposition with TM-score. It will provide a measure between 0 and 1 for the global structural similarity between the proteins and - most importantly - is length-independent.
The authors provide a web server where you can upload any two structures for comparison.
https://zhanglab.ccmb.med.umich.edu/TM-score/
Dr. Salentin has made an excellent suggestion with the TM score.
A similar metric that has been used in the Critical Assessment of protein Structure Prediction (CASP) is the Global Distance Test Total Score (GDT_TS), with a range of 0 to 100, where 100 is perfect structural agreement for the aligned residues of two proteins. Structural alignments can be carried out in various ways, such as the Mustang procedure. Here is an example of a web server that will align two pdb files and output the RMSD and the GDT_TS:
http://www.dsimb.inserm.fr/dsimb_tools/ipba/
- Badges
- Science method