FES reconstruction from metadynamics

[Frank Uhlig]

Dear all,

we are using CPMD and CP2k to perform metadynamics simulation and tried
to reconstruct the FES with the fes.sopt program.
Within CPMD we employed Gaussian tubes as history dependent biasing
potential. While comparing the reconstruction with fes.sopt and our own
program we get different results and after examinating the source code
of graph.F we concluded that the latter just adds up normal Gaussians
without narrowing. Therefore we just want to make sure that we
understood it correctly and that fes.sopt is not capable to reconstruct
FES with Gaussian tubes.
Any comments appreciated.

Best regards,

Marc and Frank