<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta content="text/html; charset=UTF-8" http-equiv="Content-Type" />
<title></title>
</head>
<body>
<p style="margin: 0px;">Hello all,</p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">As there is very little out on the web for FEP in gromacs that is up to date I thought I would throw my two cents out there for approval of my techniques (hopefully) and to help others that find this subject hard to get a significant amount of information on. Of course there are papers to be read but the specific settings and how to go about them are</p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">First if you are using 4.5+ I found that it has arguments that can be set in its mdp files:</p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">; Free energy control stuff<br />
free-energy              = yes  ; = no<br />
init-lambda              = 0.00         ; = 0<br />
delta-lambda             = 0    ; = 0<br />
foreign_lambda           =      ; =<br />
sc-alpha                 = 0.5  ; = 0<br />
sc-power                 = 1.0  ; = 0<br />
sc-sigma                 = 0.3  ; = 0.3<br />
nstdhdl                  = 10   ; = 10<br />
separate-dhdl-file       = yes  ; = yes<br />
dhdl-derivatives         = yes  ; = yes<br />
dh_hist_size             = 0    ; = 0<br />
dh_hist_spacing          = 0.1  ; = 0.1<br />
couple-moltype           = RNA_chain_A  ; =<br />
couple-lambda0           = vdw-q        ; = vdw-q<br />
couple-lambda1           = none ; = vdw-q<br />
couple-intramol          = no   ;       = no<span></span></p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">On this run I set the RNA_chain_A to be perturbation. I also set the change to be all interactions to none. To keep artifacts to a minimum I set the "sc" settings as above. (found these in the FEP tutorial online here http://www.dillgroup.ucsf.edu/group/wiki/index.php/Free_Energy:_Tutorial) I would then make several identical runs with different init_lambda values. The md or sd log file would output the average dVpot/dLambda and I would put these values in a spreadsheet vs their respective lambda values. I would then use Simpson's Rule or other to approximate the area under the curve. With this technique and lambda steps of 0.05 I calculated the FEP for the hydration of methane to be ~2.5 kcal/mol. Not perfect but in the ballpark according to the tutorial above.</p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">I found that the addition of this argument set in 4.5+ allowed for easier FEP. (ie not having to manually edit the state b for each atom in question.)</p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">If anyone else has anything so add or correct please do so. I found this very difficult to understand as the tutorial is for a different force field and my future systems were going to be much larger.</p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">Thank you,</p>
<p style="margin: 0px;"><span></span></p>
<p style="margin: 0px;"> </p>
<p style="font-family: monospace; white-space: nowrap; margin: 5px 0px 5px 0px;">TJ Mustard<br />
Email: mustardt@onid.orst.edu</p>
</body>
</html>