Fixed patches between protein variants or conformations?

[emmeburgin]

Would it be possible to use this tool to compare the changes associated with a mutation or conformation change across the same set of fixed patch definitions? For example, for quantitatively comparing the change in patch electrostatic potential illustrated in your paper in Figure 2D, without needing to rely on visually comparing the structures to identify which parts of the surface correspond to one another?

Thanks!

[fwaibl]

Hi. This is currently possible using the Python interface. To use the commandline, we'd have to implement some additional output, analogous to the hydrophobicity script. @vhoer what is your opinion on this?

E.g., if you can define your "patch" in terms of an atom selection, you can select vertices like this:

from surface_analyses.commandline_electrostatic import run_electrostatics
import mdtraj as md
import numpy as np

# This could be any selection mask matching your patch definition
patch_atoms = np.array([0, 1, 2, 3, 32])

test_traj = md.load("/home/fwaibl/Documents/code/surface_analyses/test/trastuzumab/apbs-input.pdb")
test_dx = "/home/fwaibl/Documents/code/surface_analyses/test/trastuzumab/apbs-potential.dx"
ele_output = run_electrostatics(test_traj, dx=test_dx)
surface = ele_output["surface"]
patch_vertices = np.flatnonzero(np.isin(surface["atom"], patch_atoms))
patch_vertices