On Thu, 27 Jul 2006, Geoffrey Hutchison wrote: > On Jul 27, 2006, at 2:36 AM, Tommi Hassinen wrote: > >>> Actually, what would be really nice is if the ghemical engine would >>> substitute parameters from neighbors in the periodic table instead. So if >>> I simulate Si and no parameters exist, it uses carbon but with a larger >>> vdW radius. Same thing for Se using S parameters, etc. >> >> this would work pretty well for distances and radii, but I'm afraid it >> would be too inaccurate for angle and torsion terms. :( > > You're telling me that it's more inaccurate than using hydrogen parameters? It's less inaccurate, you are right. But still I don't think it is a useful approximation. The time and effort is better spent by adding parameters for some specific elements instead (say Si for example). Regards, Tommi