Hi,
I used PRODRG ( http://davapc1.bioch.dundee.ac.uk/cgi-bin/prodrg ) to prepare topology of my ligand but recently i find out that the charges often don't come out right so we have to modify them.
Now the question is , is there other programs or web sites which prepare ligand topology in right way?
Follow the tutorial:
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin/gmx-tutorials/complex/02_topology.html
Hope it helps.
All the best.
Please see
h1ttps://www.sciencedirect.com/science/article/pii/S2352711015000059
Hi,
Do not use PRODRG. As indicated in this paper J. Chem. Inf. Model. 2010, 50, 2221–2235, the atomic charge specified by PRODRG is often unreliable. If you wish the ligand topology compatible with GROMOS forcefield, use ATB instead. According to my experience, the best choice is using acpype.py script (BMC Research Notes 2012, 5:367) to invoke antechamber (freely available in AmberTools package) to generate ligand topology for GROMACS, the topology will be based on GAFF forcefield, which is fully compatible with AMBER series of biomolecular forcefields.
Tian
Based on your force field use the program to get ligand topology. And correct it according to your need. Any of the programs can't give you 100% correct topology.
Mohammad Nazrul Islam Khan , Stéphane Abel , Tian Lu , Krishnendu Bera
Thanks every one for their kind answers.
firstly i have to apologize every one due to my incomplete question ,the force field i want to use for my ligand is GROMOS .
the answers were almost complete so thanks for them.
i'll try to use ATB web server now .
but now i have another question, is it possible and acceptable to use Gaussian for computing atom charges of ligand and then modify *.itp files charges according to Gaussian results?
Thanks.
Stéphane Abel
As i had said in last comment , using ATB will be my first alternative instead of using PRODRG as you recommended before ,
but i asked second question because i wanted to be sure that Gaussian would be helpful in same cases.
But of course your comment and recommendation is much better and easier and i'll use that.
anyway thanks alot
Good luck!
Using Gaussian or other code (such as my wavefunction analysis code Multiwfn) to compute atomic charges and write them into molecule .itp file is completely acceptable. But you need to make sure that the way of calculating atomic charge you used is compatible with the forcefield. Different forcefields use different ideas to derive atomic charges. For example, most modern AMBER series of forcefield use RESP atomic charges (the cheaper AM1-BCC atomic charges are also usable). More information can be found in original paper of the forcefields. Regardless of type of forcefield, the key requirement on the method of deriving atomic charges is that the resulting atomic charges must be able to well reproduce molecular electrostatic potential (ESP) around molecular van der Waals surface, otherwise intermolculear electrostatic interaction will be unable to be described well. (I did a systematical test for most popular atomic charge calculation methods, see Table 5 and 6 in doi: 10.3866/PKU.WHXB2012281, methods with relatively low RRMSE is in general suitable for MD simulation purpose)
Regarding GROMOS, charge group concept is utilized in assigning atomic charges, see related papers about ATB server for discussion. The atomic charges provided by ATB server is in line with this philosophy. It is difficult to obtain atomic charges in principle 100% compatible with GROMOS by yourself using other codes. However, in practical situations, it is still reasonable to use ESP fitting methods (e.g. CHELPG and MK) to obtain atomic charges for rigid molecules and use RESP charges for flexible molecules when simulating them in the framework of GROMOS FF.
Use ATB. Even the GROMACS developers point towards the drawbacks of using PRODRG.
ATB works fine on my systems.
- Badges
- Science topic