Interpretation of WATERSWAP results

[yusras...@gmail.com]

Dear Christopher,

I obtained the following results for four of my CYP3A4-inhibitor complexes

Bennetts = 46.06587822472148 +/- 0.050242921483837605 kcal mol-1
FEP = 43.60892216376602 +/- 14.145502424176618 kcal mol-1
TI = 43.387058478162935 +/- 3.1260639789048312 kcal mol-1 (quadrature = 43.31624975352257 kcal mol-1)

Free energies
Bennetts = 20.730281635658752 +/- 0.06945016576387397 kcal mol-1
FEP = 20.15382638967887 +/- 18.103366149323154 kcal mol-1
TI = 21.05752253189108 +/- 2.963616632777857 kcal mol-1 (quadrature = 20.619743108428725 kcal mol-1)

Free energies
Bennetts = 25.47837048568636 +/- 0.0678113389289521 kcal mol-1
FEP = 21.918017573997563 +/- 14.715419466643336 kcal mol-1
TI = 25.23500933356021 +/- 2.1670812485134157 kcal mol-1 (quadrature = 25.5934079213815 kcal mol-1)

Free energies
Bennetts = 17.75348435530676 +/- 0.10075198136989356 kcal mol-1
FEP = 15.634026210223313 +/- 30.942109019810246 kcal mol-1
TI = 15.950507435892883 +/- 3.007881011717288 kcal mol-1 (quadrature = 14.609768634795888 kcal mol-1)

where as in case of remaining CYP3A4-inhibitor complex I obtained the following free energies

Free energies
Bennetts = 12.461753572360125 +/- 0.07536451806129024 kcal mol-1
FEP = 9.670326455532809 +/- 38.14067670375319 kcal mol-1
TI = -353.10221921959027 +/- 643.8608364897641 kcal mol-1 (quadrature = -380.60641614924737 kcal mol-1)

Since I understand these energies should be in similar range and may not differ considerably but what might be the reason for this difference for the aforementioned case? and what should be done to obtain correct results.


Regards,
Yusra

[Christopher Woods]

  Hi Yusra,

How many iterations did you perform for each inhibitor complex? The calculations give four estimates for the free energy, e.g. for 

Bennetts = 46.06587822472148 +/- 0.050242921483837605 kcal mol-1
FEP = 43.60892216376602 +/- 14.145502424176618 kcal mol-1
TI = 43.387058478162935 +/- 3.1260639789048312 kcal mol-1 (quadrature = 43.31624975352257 kcal mol-1)

you have four estimates; 46.1, 43.6, 43.4 and 43.3 kcal mol-1. The level of agreement gives an idea of the convergence of the result. In this case, the binding free energy is around -43 to -46 kcal mol-1, which is a little strong, and the level of disagreement between the numbers is a little high (I prefer them all to agree to within 1.0 kcal mol-1).

You second binding free energy does show good agreement, i.e. -20.2 to -21.1 kcal mol-1, which looks a lot more reasonable.

Your third binding free energy is -14.6 to -17.8 kcal mol-1. which looks like a reasonable value, but the level of disagreement is too high.

The last value is -9.8 to +380 kcal mol-1, which suggests that either something has gone wrong or the simulation needs many more iterations to converge.

So, to obtain better results, you need to run the simulations that show poor convergence for longer, i.e. more iterations. The last result in particular looks suspect, and you should take a look at the PDB files in the "output" directory to see if everything looks ok with the structure. The very large values indicate that there are some clashes of molecules, i.e. maybe the ligand or water is squashed too tightly with the protein.

  Best wishes,

  Christopher

--
You received this message because you are subscribed to the Google Groups "Sire Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sire-users+unsubscribe@googlegroups.com.
To post to this group, send email to sire-...@googlegroups.com.
Visit this group at https://groups.google.com/group/sire-users.
For more options, visit https://groups.google.com/d/optout.

--

---------------------------------------------------------
Christopher Woods
+44 (0) 7786 264562
http://chryswoods.com

[yusras...@gmail.com]

Hi Christopher,

I ran my waterswap calculations for 1000 iterations. Can you please explain a little more how can I improve my results. What should be observed while visualizing the pdb files generated in the output folder. Do I need to inspect all the pdbs generated in the output folder?

Best wishes

Regards,

Yusra Sajid 

To unsubscribe from this group and stop receiving emails from it, send an email to sire-users+...@googlegroups.com.

To post to this group, send email to sire-...@googlegroups.com.
Visit this group at https://groups.google.com/group/sire-users.
For more options, visit https://groups.google.com/d/optout.

[Christopher Woods]

  Hi Yusra,

Running for 1000 iterations should have converged the free energies, although the results indicate otherwise. When looking at the PDB files you are trying to see whether there are any overlaps between the ligand, water cluster or protein, or if there are weird or unusual conformations of the ligand or protein.

Yes, you should look at all of the bound_*.pdb files that are in the output directory. You can do this as a movie in VMD by typing

vmd output/bound_*0.005*.pdb

(the above command will show the lambda=0.005 trajectory, i.e. when the ligand is bound to the protein)

or

vmd output/bound_*0.995*.pdb

(the above command will show the lambda=0.995 trajectory, i.e. when the water cluster is bound to the protein).

You can visualise the swapped water cluster by highlighting the molecules with "resname SWP". The swapped water cluster will overlap with the ligand, as this is being swapped with the ligand.

I would particularly look at the first ligand (the one with free energies of around 40 kcal mol-1), as these values are very high, which indicates that the swap water cluster did not bind as well to the protein as the ligand. Looking at the structure in VMD may help identify why (maybe the swap water cluster was too big or too small to fit into the binding site?).

I would also look at the last ligand (the one showing free energies between -9.8 to +380 kcal mol-1), as this implies that the calculation did not converge at all. This suggests that something went wrong, i.e. the ligand was badly overlapping with the protein, or the swap water cluster was badly overlapping with the protein.

  Best wishes,

  Christopher

To unsubscribe from this group and stop receiving emails from it, send an email to sire-users+unsubscribe@googlegroups.com.

To post to this group, send email to sire-...@googlegroups.com.
Visit this group at https://groups.google.com/group/sire-users.
For more options, visit https://groups.google.com/d/optout.