Force constants with quartic term

[Alexander Fokas]

Hello,

I am aware that there is an option for specifying user-defined spring constants (scalar quantities). However, the springs will still be set up by connecting the atoms together (thus giving vector forces between the atoms), assuming F=kx.

I am hoping to compute the forces using a quartic term, thus F=kx will no longer hold. I am wandering if this is possible within Bio3d, or whether I should turn to Matlab alongside Bio3d to compute the forces.

Many thanks,

Alexander

[Xinqiu Yao]

Hi,

I think bio3d internally calculates normal modes based on the elastic network model (assuming linear force for each pair of interacting residues), and so it is not straightforward if you are actually considering other form of energy function. Yes, if you know how to calculate forces (and also the Hessian) with Matlab, it should be easy by passing the corresponding Hessian to nma.pdb(). Alternatively, you may try hacking the build.hessian() function to match your purpose... Am I correct, Lars?

[Lars Skjærven]

Jepp - the function build.hessian assumes an harmonic Hookean potential. You'll have to make your build hessian function (potentially in R to incorporate it to bio3d? you can surely also do it in matlab, and export the matrix to R). When you have your hessian matrix then you can use a set of bio3d functions to build your nma object as Xinqiu mentioned.

[Alexander Fokas]

I think that is how I will proceed!

Cheers