3. Protein-protein interactions

This notebooks shows how to compute a fingerprint for protein-protein interactions.

Here we will investigate the interactions in a G-protein coupled receptor (GPCR) between a particular helix (called TM3) and the rest of the protein.

This can obviously be applied to proteins that don’t belong to the same chain/segment, as long as you can figure out an appropriate MDAnalysis selection

There is also an example at the end of this tutorial for generating an IFP of PPI without considering the backbone.

import MDAnalysis as mda
import prolif as plf

# load traj
u = mda.Universe(plf.datafiles.TOP, plf.datafiles.TRAJ)
tm3 = u.select_atoms("resid 119:152")
prot = u.select_atoms("protein and not group tm3", tm3=tm3)
tm3, prot

(<AtomGroup with 531 atoms>, <AtomGroup with 4457 atoms>)

# prot-prot interactions
fp = plf.Fingerprint(["HBDonor", "HBAcceptor", "PiStacking", "PiCation", "CationPi", "Anionic", "Cationic"])
fp.run(u.trajectory[::10], tm3, prot)

<prolif.fingerprint.Fingerprint: 7 interactions: ['HBDonor', 'HBAcceptor', 'Cationic', 'Anionic', 'PiCation', 'CationPi', 'PiStacking'] at 0x7f79207b0ca0>

df = fp.to_dataframe()
df.head()

ligand	GLN119.A										...	ALA150.A		ILE151.A			THR152.A
protein	GLN189.A	ALA190.A	ALA192.A		GLU194.A		VAL196.A		SER197.A		...	ALA154.A	ARG238.A	ALA154.A	ARG238.A	ASN245.A	ASP153.A		GLU234.A	ARG238.A	LYS241.A
interaction	HBAcceptor	HBDonor	HBDonor	HBAcceptor	HBDonor	HBAcceptor	HBDonor	HBAcceptor	HBDonor	HBAcceptor	...	HBAcceptor	HBAcceptor	HBAcceptor	HBAcceptor	HBAcceptor	HBDonor	HBAcceptor	HBDonor	HBAcceptor	HBAcceptor
Frame
0	False	False	True	False	False	True	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False
10	False	False	False	False	False	False	False	False	False	False	...	False	False	False	True	False	False	False	False	False	False
20	False	False	False	False	False	False	False	False	False	True	...	False	False	False	True	False	False	True	False	True	False
30	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False
40	False	False	False	False	True	False	False	True	False	False	...	False	False	True	True	False	False	False	False	False	False

5 rows × 55 columns

# show interactions for a specific ligand residue
df.xs("ARG147.A", level="ligand", axis=1).head(5)

protein	ALA84.A	GLU309.B		THR313.B
interaction	HBDonor	HBDonor	Cationic	HBDonor
Frame
0	False	False	False	False
10	False	False	False	False
20	True	False	False	False
30	False	False	False	False
40	False	True	True	False

# same for a protein residue
df.xs("GLU309.B", level="protein", axis=1).head(5)

ligand	ARG147.A
interaction	HBDonor	Cationic
Frame
0	False	False
10	False	False
20	False	False
30	False	False
40	True	True

# display a specific type of interaction
df.xs("Cationic", level="interaction", axis=1).head(5)

ligand	ARG147.A
protein	GLU309.B
Frame
0	False
10	False
20	False
30	False
40	True

# calculate the occurence of each interaction on the trajectory
occ = df.mean()
# restrict to the frequent ones
occ.loc[occ > 0.3]

ligand    protein   interaction
CYS122.A  GLY118.A  HBDonor        0.84
ASP123.A  LYS191.A  Anionic        0.32
TRP125.A  VAL102.A  HBDonor        0.60
          TYR109.A  PiStacking     0.48
          TRP115.A  PiStacking     0.80
SER127.A  SER181.A  HBDonor        0.32
ASP129.A  TYR359.B  HBAcceptor     0.96
HSD139.A  TRP174.A  PiStacking     0.36
ASP146.A  TYR157.A  HBAcceptor     0.88
          ARG161.A  HBAcceptor     0.48
                    Anionic        0.36
ARG147.A  GLU309.B  HBDonor        0.44
                    Cationic       0.44
TRP149.A  ASP153.A  HBAcceptor     0.36
dtype: float64

# regroup all interactions together and do the same
g = (df.groupby(level=["ligand", "protein"], axis=1)
       .sum()
       .astype(bool)
       .mean())
g.loc[g > 0.3]

ligand    protein
ARG147.A  GLU309.B    0.44
ASP123.A  LYS191.A    0.32
ASP129.A  TYR359.B    0.96
ASP146.A  ARG161.A    0.48
          TYR157.A    0.88
CYS122.A  GLY118.A    0.84
HSD139.A  TRP174.A    0.36
SER127.A  SER181.A    0.32
TRP125.A  TRP115.A    0.80
          TYR109.A    0.48
          VAL102.A    0.60
TRP149.A  ASP153.A    0.36
dtype: float64

3.1. Ignoring backbone interactions

In some cases, you might want to dismiss backbone interactions. While it might be tempting to just modify the MDAnalysis selection with "protein and not backbone", this won’t work as expected and will lead to adding a charges where the backbone was bonding with the sidechain.

However there is a temporary workaround (which will be directly included in the code in the near future):

from rdkit import Chem
from rdkit.Chem import AllChem
from tqdm.auto import tqdm

# remove backbone
backbone = Chem.MolFromSmarts("[C^2](=O)-[C;X4](-[H])-[N;+0]")
fix_h = Chem.MolFromSmarts("[H&D0]")

def remove_backbone(atomgroup):
    mol = plf.Molecule.from_mda(atomgroup)
    mol = AllChem.DeleteSubstructs(mol, backbone)
    mol = AllChem.DeleteSubstructs(mol, fix_h)
    return plf.Molecule(mol)

# generate IFP
ifp = []
for ts in tqdm(u.trajectory[::10]):
    tm3_mol = remove_backbone(tm3)
    prot_mol = remove_backbone(prot)
    data = fp.generate(tm3_mol, prot_mol)
    data["Frame"] = ts.frame
    ifp.append(data)
df = plf.to_dataframe(ifp, fp.interactions.keys())
df.head()

ligand	GLN119.A		CYS122.A	ASP123.A					PHE124.A	TRP125.A	...	ASP146.A			ARG147.A			TYR148.A	TRP149.A	THR152.A
protein	GLN189.A	SER197.A	LYS191.A	ARG188.A		GLN189.A	LYS191.A		PHE185.A	SER106.A	...	TYR157.A	ARG161.A		GLU309.B		THR313.B	ARG230.A	TYR157.A	ASP153.A	GLU234.A
interaction	HBAcceptor	HBAcceptor	Anionic	HBAcceptor	Anionic	HBAcceptor	HBAcceptor	Anionic	PiStacking	HBDonor	...	HBAcceptor	HBAcceptor	Anionic	HBDonor	Cationic	HBDonor	HBAcceptor	PiStacking	HBDonor	HBDonor
Frame
0	False	False	False	False	False	False	False	False	False	True	...	True	False	False	False	False	False	False	False	False	False
10	False	False	False	False	False	False	False	False	False	False	...	True	False	False	False	False	False	False	True	False	False
20	False	True	False	False	False	False	True	True	False	False	...	True	False	False	False	False	False	False	True	False	False
30	False	False	False	True	True	False	False	False	False	False	...	True	True	True	False	False	False	False	False	False	False
40	False	False	False	False	False	False	True	True	False	True	...	True	False	False	True	True	False	False	False	False	False

5 rows × 28 columns