primaryodors / primarydock

PrimaryOdors.org molecular docker.
Other
6 stars 4 forks source link

Ionic interactions not working. #20

Closed electronicsbyjulie closed 2 years ago

electronicsbyjulie commented 2 years ago

Using protonated cadaverine with the formula [NH3+]CCCCC[NH3+] and TAAR8.

The result is zero ionic bonding.

Total ionic: -0
Total H-bond: -18.3215
Total pi stack: -0
Total polar-pi and cation-pi: -0.427101
Total metal coordination: -0
Total van der Waals: -11.236
Total: -21.4182
electronicsbyjulie commented 2 years ago

Not quite.

Pose: 3
Node: 0
# Binding energies
BENERG:
...
SER108: -23.379
ASP111: -0
VAL112: -14.7409
...
ILE200: -10.0142
ASP201: -0
LEU203: -1.88846
...
Total ionic: -663.586
Total H-bond: -2.33858
Total pi stack: -0
Total polar-pi and cation-pi: -1.67147
Total metal coordination: -0
Total van der Waals: -8.74317
Total: -85.905

ASP111 and ASP201 are coordinated beautifully to the ligand, but in the results they both have a binding energy of zero. They should have the strongest negative energies of the list. TAAR8_ASPs

Also, why is the total ionic energy -663.586??? There aren't enough residues in the result with negative energies to account for that. It looks like the ions aren't shielding properly, and are registering high energies even when physically far apart. There is, however, a large number of positive results, which is not good since this means more van der Waals collisions than the real molecules would allow.

electronicsbyjulie commented 2 years ago

At some point this was fixed.

ASP111: -51.3797
...
ASP201: -3.7756
...
Total ionic: -73.6867
Total H-bond: -4.11537
Total pi stack: -0
Total polar-pi and cation-pi: -0.237792
Total metal coordination: -0
Total van der Waals: -7.19147
Total: -39.4692

It is now only coordinating the ligand to ASP111 and not to ASP201. This is because there is only one "best bind" heavy atom of the ligand and only one alignment point inside the binding pocket. In theory, there can be up to three "best bind" coordinations without having to bend the ligand.

electronicsbyjulie commented 2 years ago

30.