compare displacement with dccm

[jay]

I have a graph giving cRMS displacement based on an all-atom NMA of a protein (not generated from Bio3D). Now for the same protein, I have a dccm plot from a NMA (slow modes with freq. less than 30cm-1 only) based on C-alpha coordinates (generated from Bio3D) giving the coupled regions in terms of motion. In what way are the 2 results complementary ?

While the 1st graph is a plot of cRMS displacement on the residue indices, the dccm plot gives me correlated regions.

The problem was that there was not enough memory to carry out aanma in the standalone Bio3D so I did it on a server of another NMA tool. But there are no downstream analysis facility available in that particular tool. So performed the C-alpha NMA (which was feasible on my laptop) and plotted DCCM.

Please let me know your thoughts on the difference between the info. specified by the 2 plots ?

Will it be meaningful to include both the plots in a paper ?

[Xinqiu Yao]

Hi,

It is not very clear to me what you are trying to get answered here. aaNMA and C-alpha NMA are different models and which one to use is largely dependent on objectives of study. In our test, aaNMA works better for cases where detailed information about residue-residue correlations is important (such as correlation network analysis). However, with respect to predictions of overall atomic fluctuations and covariance, both models work equally well.

I would recommend focus on analyzing the results first - It usually becomes clearer which model is more suitable after interpreting their output (and this needs your own expertise about the system you are working on that others can't help on behalf of you). Normally, you only need one model and stick to it. However, if both models give similar results, you might consider including both (one in main text and the other in supplemental for example) to show the robustness of your results.

Hope it helps...

[Barry Grant]

Or if the two models give very different results then this should be investigated further to find out the reason - this could actually be very informative ;-)

[Barry Grant]

Q&A