[gmx-developers] DNA is one moltype or two moltype?

[toannt]

Hi,

Thanks. I'm defining two molecules types, each type have only one 
molecule in the simulation. When I said custom bonds, I was referring to 
the hydrogen bonds between the two strands. I'd like to understand how 
these bonds are handled by Gromacs in an atomistic simulation? If I have 
two strands at some distance apart, do they naturally come to each other 
to form the dsDNA?

By the way, I am doing a grand canonical simulation, but I'd like to 
reuse libgromacs to avoid reinventing the wheel with some functions.

Toan

Vào ngày 14-10-2015 17:56, Berk Hess viết:
> Hi,
>
> There is no direct link between "real" molecules and molecule types.
> The only restriction is that you can only define chemical bonds
> between atoms within the same molecule type. Thus you can have two 
> DNA
> strands in one or in two separate molecule types.
> I don't understand what custom-bonds you are referring to. Usually
> DNA strands are not chemically bonded to each other.
>
> Cheers,
>
> Berk
>
> On 10/14/2015 11:49 AM, toannt wrote:
>>
>> Hi all
>>
>> I've just started to look at libgromacs for a custom simulation. I'm 
>> studying the gmx data structures to initialize them directly. I wonder 
>> if molecules such as DNA is considered by gromacs to be two molecule 
>> types (one for each strand) or is only one molecule type with some 
>> custom-bonds between the two strands?
>>
>> Thanks
>>
>> Toan