< 7th term

Protein surface

Last update on the 13th of December, 2019

Here we study surface area of transhydrogenase NMR model (2BRU, fig. 1) and chain interface surface.

Fig. 1. 2BRU molecular surface and cartoon representation. Green: chain A, Blue: chain B, Magenta: chain C.

Surface area

2BRU consists of 10 models with 3 chains. MS area and SAS area were calculated with PyMol get_area for each model (fig. 2).

Fig. 2. 2BRU MS and SAS areas of every model in the record.

MS and SAS areas are generally correlated (Pearson's r=0.765, Kendall's τ=0.689). Area differences between models might be explained with little conformation changes. In addition, MS area values have lower standard deviation than SAS area values (98Å and 419Å, respectively) thus indicating more dramatic SAS changes over conformations.

Chain interface

Chain interface residues were estimated with interfaceResidues.py script from PyMol Wiki for each pair of chains.

Fig. 3. Chain interface residues. A: surfaces are not split between chains. B: surfaces are split between chains. Interface residues represented as sticks, interface surface is shown.

Discovered residues reveal rich interface between chains (fig. 3). PyMol builds unified surface in case chains comprise single object. In case chains are split in multiple objects (with PyMol extract command), interface surfaces intersect each other as it should be defined by the algorithm of interface discovery.

Fig. 4. Chain interface residues in-depth. Left: surfaces are not split between chains. Right: surfaces are split between chains. Interface residues represented as sticks, interface surface is shown.

However, it is not the case in general (fig. 4): not all surfaces of interacting residues intersect each other. That's definitely a bug. Nonetheless, interacting surfaces complement each other.