Reweighting in well tempered simulation

[Sk Habibullah]

Hello all,

In the Lugano tutorial: Metadynamics simulations with PLUMED

Reweighting section, 

after doing plumed driver --ixtc traj_comp.xtc --plumed plumed.dat --kt 2.5

COLVAR file is obtained.

#! FIELDS time phi psi metad.bias #! SET min_phi -pi #! SET max_phi pi #! SET min_psi -pi #! SET max_psi pi 0.000000 -1.497988 0.273498 110.625670 1.000000 -1.449714 0.576594 110.873141 2.000000 -1.209587 0.831417 109.742353 3.000000 -1.475975 1.279726 110.752327

The last column will give as, in energy units, the logarithm of the weight of each frame. You can easily obtain the weight of each frame using the expression w∝exp(V(s)/kBT). You might want to read the COLVAR file in python and compute a weighted histogram.

this column 2 and 3 are biased data, right?

but if metad.bias is V(s) then exp(V(s)/kBT) will be very high value (considering V(s) is in kJ/mol and kBT as 8.314*300*0.001 kJ/mol). This weights I have to divide with phi and psi in column 2 and 3 to get the unbiased data (eleminating bias) right? 

what is the expression of w , here it is given as ∝exp(V(s)/kBT). How can I compute w and what should I do with w to get unbias data?

Please let me know.

Thank you,

Sk Habibullah

IISc Bangalore

[Sk Habibullah]

Could someone tell some hints on that issue? 

Thanks

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/d8ddc107-1425-4d53-b2a9-d7bbad1eabf8n%40googlegroups.com.

[Giovanni Bussi]

Hi,

the usual trick is to first remove the maximum value from V

V-=np.max(V)

Then you do the exponential (which will be equal to 1 at most, since max(V) is zero now), and you normalize

Giovanni

To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/CAGY7OeJhnXY3c1boY4NzRdoF%3D%3DRc6xLv8B8qhXNuWJzSvFfJ_w%40mail.gmail.com.

[Sk Habibullah]

Okay thank you, that will give me w. Then should I divide the w with biased values of the CVs to get the unbiased values for each snapshot?

To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/CAPLm8ZLXFHNXAyHFPpypNZp2ymohE72cDGEB2XpRoiLkDkrhMQ%40mail.gmail.com.