Vibrational entropy and the effect of mutations

Among coarse-grained normal mode analysis methods, ENCoM has the unique ability of being sensitive to the chemical nature of the amino acids, nucleic acids and/or ligands.

This allows ENCoM to compute the change in vibrational entropy between WT and mutant forms of the same protein. In the nrgten_examples repository, the FimA protein from E. coli (PDB id 6R74) has had position 103 mutated from an isoleucine to a tyrosine using MODELLER (doi.org/10.1002/cpbi.3). The difference in vibrational entropy between the WT and mutant can be computed in this way:

from nrgten.encom import ENCoM

wt = ENCoM("6r74.pdb")
tyr103 = ENCoM("6r74_TYR103.pdb")
wt_entropy = wt.compute_vib_entropy(beta=1)
tyr103_entropy = tyr103.compute_vib_entropy(beta=1)
diff_entropy = wt_entropy - tyr103_entropy

The diff_entropy variable above contains the difference in vibrational entropy between the WT and mutant forms of the protein. If we print its value using:

print(diff_entropy)

We see that it is around -0.02, which means that the mutant is more flexible than the WT.

Note

The beta parameter, which is the Boltzmann scaling factor, is set to 1 in the above examples because preliminary work has shown this value to give the best results when trying to predict experimental measurements of ΔΔG.

Note

Since the ENCoM model is pseudo-physical, the vibrational entropy value does not have definite units (but it still represents a measurement of energy per temperature). It is however possible to select a beta parameter that gives a close match to experimental data, or to use the vibrational entropy value in a linear predictor which will match experimental data.