MP2 is great, but MP2 gradients are expensive and I'm not sure how large your systems are.

If you can afford to use MP2, then you can probably afford to use better DFT functionals than B3LYP - I would look into a good benchmark study, for example;

http://pubs.rsc.org/en/content/articlehtml/2017/cp/c7cp04913g

And pick one of the recommended functionals from there. They even have a breakdown of performance in terms of different chemical properties.

I would probably go with something like PW6B95-D3 and then perform singlepoints with a double-hybrid functional to get the MP2 correlation if nescessary.

You can use B3LYP in combination with a good basis set or M06-2X with good basis set. cc-pV(n+d)Z or 6-311++G(3d2f,2df,2p)

MP2 is great, but MP2 gradients are expensive and I'm not sure how large your systems are.

If you can afford to use MP2, then you can probably afford to use better DFT functionals than B3LYP - I would look into a good benchmark study, for example;

http://pubs.rsc.org/en/content/articlehtml/2017/cp/c7cp04913g

And pick one of the recommended functionals from there. They even have a breakdown of performance in terms of different chemical properties.

I would probably go with something like PW6B95-D3 and then perform singlepoints with a double-hybrid functional to get the MP2 correlation if nescessary.

Mrs/Miss. Das,

A meaningful analysis means a comparative analysis between the methods. Because of the type of the level depends on the atoms of interest and the type of the interactions. Noncovalent interactions means classical H-bonds, short contacts, ion-pi-interactions, etc.

Dear all,

Thank you very much for the responses. I will surely look into all these methods before proceeding with my work.

see this ppt at last 4 slides