Any molecular dynamics run consists of four important steps.
When we packed the system using Packmol, we specified a tolerance and based on that the system was created. But we need our system to be stable
or to be at minimum energy. This is carried out in the minimization step using some of the well known algorithms like the conjugate gradient method
or the steepest descent method. This step is carried out at 0K (this means that we won't be required to handle the 3N velocity information).
But our simulations aren't carried out at 0K, as such our next logical step is to heat it to the system specified temperature (some people heat it even further
and then cool it back, in a cycle, this is called annealing, I'll cover that later). This logical step is named "Heating".
After heating, we have our system at the native step,
up till now no molecular dynamics run has been performed, meaning we haven't utilized the information in the parm file. The actual molecular dynamics
starts here. However, we cannot expect to get useful information the moment simulation starts, instead we allow the engine to heat up a beat before we calculate the milage
(God bless the analogy!). This step is named "Equilibration". Finally, equilibration step is followed bt the Production run. These are the "Molecular Dynamics run". The values obtained here are used to
obtain useful information about the system and to analyze the results obtained from our calculations. If you think it is a rather illogical thing to carry out calculations in computer
which are so different from the real world, well, the Noble Prize in 2013 went for MD simulations. Take that!!
Here is a sample configuration file for Minimization run, copied in part from NAMD manual, and in part from Green solvent lab member's cf. Create a new folder and save this file. Look how I have kept all the psf, pdb files together in a different foler named "common". Running the script won't be an issue, what is important is to understand the script. I have tried to explain a few commands in the comments in the file itself, however, this is temporary and you need to have a profound understanding of the text. Have a look here. Go through each of the step diligently and arduously taking pain to understand what each line stands for.
Here is a sample configuration file for Heating run, copied in part from NAMD manual,
and in part from Green solvent lab member's cf. Save this file. Running the script won't be an issue, what is important is to understand the script.
Have a look here. Go through each of the step diligently and arduously taking pain to understand what each line stands for.
In case you wish to perform annealing simulations, you need to create different folders. I will keep this tutorial based on annealing run; this way those who
wish to perform annealing can follow suit, others can edit this text file as per their need.
We will heat the system 1K in 200 steps (50K per 10000 steps!). Once the temperature reaches a top value of 350K, it will be cooled back to
300K, then again heat it to 350K, and the cycle will go on for 3 loops. There is no need to save all these files, but we will. As such, we will create
three different folders. More on these are discussed in next section. For now, have a look at the script and move on.
Here is a sample configuration file for Equilibration run, copied in part from NAMD manual,
and in part from Green solvent lab member's cf. Save this file. Running the script won't be an issue, what is important is to understand the script.
Have a look here. Go through each of the step diligently and arduously taking pain to understand what each line stands for.
We will have an equilibration run of 3ns (1500000 steps). Have a good look at the configuration file, it is a complete MD file.
A good practice is to reduce the freq of different output files, the output files of equilibration run are generally deemed useless.
Here is a sample configuration file for Production run, copied in part from NAMD manual,
and in part from Green solvent lab member's cf. Save this file. Running the script won't be an issue, what is important is to understand the script.
Have a look here. Go through each of the step diligently and arduously taking pain to understand what each line stands for.
To carry out the static property calculations, like energy calculation or system density calculation, we
carry forward our equilibration NPT run for desired steps. We ensure that the output frequency of different files are not very high.
The more output files we have, the better and more refined calculations we can obtain (depending on our system capabilities obviously).
This step will be carried out for 4ns (2000000 steps)
Keep a check on the .xsc (or the .xst) file, it will give you box lengths. We will use that to obtain the density of the system. The calculation is simple.
We have 500 molecules of ethanol in a box of volume V (say 150 angstrom cube). Next, we find the mass of those 500 molecules (you know this part, right?).
Divide this mass by the volume of the system and we have our density. The experimental value of the density of ethanol at 300K is 0.81g/cc.
After the simulation ends, you can check what value you obtain!
Previous Page Main Page Next Page