<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.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;">Moeed,</p>
<p style="margin: 0px;"><span><br />
</span></p>
<p style="margin: 0px;"><span>Sorry for the late response, and yes I did need to include more information in my answer.</span></p>
<p style="margin: 0px;"><br />
<span> </span></p>
<p style="margin: 0px;"><span>1) If your polymers are physically close together in reality you might want to have multiple polymers in a single FEP job.</span></p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">I run single ligands in proteins, so this has never been an issue. You may want to run both and see what the differences in energy are. If there are none you may be able to do longer md/sd runs with a smaller one polymer system than you could with multiple polymers in the same time.</p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">In other words I have no idea what your system will need. Sorry.</p>
<p style="margin: 0px;"><br />
<span> </span></p>
<p style="margin: 0px;"><span>2) How many windows and how long are the md/sd runs for each window?</span></p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">I personally found that 21 windows works well for my systems. This however will vary by the dV/dlambda "graph". If and when your dV/dlambda is changing more (ddV/dlambda) you will need more windows to get a more accurate energy. Most systems change the most at the beginning and end of lambda values. I run a lambda every 0.05, but my systems are huge! (>100,000 atoms)</p>
<p style="margin: 0px;"><br />
<span> </span></p>
<p style="margin: 0px;"><span>3) Bennett acceptance ratio is an extra step in setup but makes getting your data much easier to interpret. Once you are done with all your windows you run g_bar and it will print out on your screen the energy from each window and the total energy in kJ/mol with a standard deviation.<br />
</span></p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">This can be found in the manual and also in the gmx_user archives. This entailes making a .mdp file for every step (but you probably do this already) and setting the foreign lambdas correctly.</p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">ie for lambda values: 0.00, 0.25, 0.50, 0.75 and 1.00</p>
<p style="margin: 0px;"> Your foreign lambda values for the 0.00 mdp file would be:  0.25</p>
<p style="margin: 0px;"> Your foreign lambda values for the 0.25 mdp file would be: 0.00 0.50</p>
<p style="margin: 0px;"> Your foreign lambda values for the 0.50 mdp file would be: 0.25 0.75</p>
<p style="margin: 0px;"> Your foreign lambda values for the 0.75 mdp file would be: 0.50 1.00</p>
<p style="margin: 0px;"> Your foreign lambda values for the 0.50 mdp file would be: 0.75 </p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">This of course is a small number of lambda values you will want to use more.</p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">Please email me with questions if you need more.</p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">Thank you,</p>
<p style="margin: 0px;">TJ Mustard</p>
<p style="margin: 0px;">Email: mustardt@onid.orst.edu</p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">PS more below</p>
<p style="margin: 0px;"><span><br />
</span></p>
<div style="margin: 5px 0px 5px 0px;">
<br />
On February 14, 2011 at 8:54 PM Moeed <lecielll@googlemail.com> wrote:<br />
<br />
<blockquote type="cite" style="margin-left: 0px; padding-left: 10px; border-left: solid 1px blue;">
<div>
Hello,
</div>
<div>
 
</div>
<div>
Sorry but it seems you wanted to answer my question on gmx list but I dont see anything!...
</div>
<div>
 
</div>
<div>
<br />
<br />
--<br />
Moeed Shahamat<br />
Graduate Student (Materials Modeling Research Group)<br />
McGill University- Department of Chemical Engineering<br />
Montreal, Quebec H3A 2B2, Canada<br />
Web:<a href="http://mmrg.chemeng.mcgill.ca/pages/current-group-members/moeed-shahamat.php">http://mmrg.chemeng.mcgill.ca/pages/current-group-members/moeed-shahamat.php</a><br />
Web:<a href="http://mmrg.chemeng.mcgill.ca/">http://mmrg.chemeng.mcgill.ca/</a>
</div>
</blockquote>
</div>
<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>
<p style="font-family: monospace; white-space: nowrap; margin: 5px 0px 5px 0px;"> </p>
<p style="font-family: monospace; white-space: nowrap; margin: 5px 0px 5px 0px;"> </p>
<blockquote>
<p style="margin: 0px;">Dear experts,<br />
<br />
I am going to do solvation FE of polymer (polyethylene) in a hydrocarbon solvent. I have prepared a system consisting of 4 polymer chains and 480 hexane molecules with the actual density of polymer solution (~ 0.5 g/cm3).<br />
<br />
1- For such a study I dont know how many polymers I need to have in my system. If FE can be done with only one chain, am I making system bigger in vain? Does this matter affect the accuracy of results?</p>
</blockquote>
<div style="margin: 5px 0px;">
<blockquote type="cite" style="margin-left: 0px; padding-left: 10px; border-left: 1px solid blue;">
2- I have switched off electrostatics so I am using<br />
<br />
free_energy          =   yes<br />
init_lambda          =   0 <br />
delta_lambda         =   0<br />
sc_alpha             =   0.5<br />
sc-power             =   1<br />
sc_sigma             =   0.3<br />
couple-lambda0       =   vdw<br />
couple-lambda1       =   none<br />
couple-intramol      =   no<br />
<br />
In David Mobley's turorial the last three lines are not included. I wanted to know if I am to run say 10 simulations for different lambda, what purpose does the last three lines serve in 4.0.7  ? I got very close values in that tutorial without these settings. ( I know what these lines mean, just curious how these three lines affect the results in 4 X +).<br />
<br />
<br />
</blockquote>
</div>
<p style="margin: 0px;">This old tutorial is for an old gromacs which needs you to set the state a and state b. In the newer gromacs releases you set the state a, and gromacs will go to state b without you setting each atom. This considerably saves time making .itp files. All you need is a methane .itp file (got mine from antechamber) and then you set couple-lambda0 and couple-lambda1 to what you are looking for.</p>
<p style="margin: 0px;"> </p>
<p style="margin: 0px;">I personally go from couple-lambda0 = vdw-q  to couple-lambda1 = none.</p>
<div style="margin: 5px 0px;">
<blockquote type="cite" style="margin-left: 0px; padding-left: 10px; border-left: 1px solid blue;">
Please let me know your comments/point of view about the system and setting I am using.<br />
<br />
Thanks<br />
Moeed<br />
</blockquote>
</div>
<p style="margin: 0px;"> </p>
<p style="font-family: monospace; white-space: nowrap; margin: 5px 0px;">TJ Mustard<br />
Email: mustardt@onid.orst.edu</p>
<br />
<br />
</body>
</html>