I agree that 0.4 mM is very low. So if you calculate the volume of your simualtion box and calculate the number of ions you should introduce probbably (depending on the size of your system) you'll find out a very small number.

In principle, however, there are 2 points that deserve comments:

1. the initial size of your box will likely change during the MD, since yousually you may want to run an NPT simulation. So the Ca2+ conc. should be calculated with respect to the "final equilibrated" box, otherwise you'll have a higher Ca2+ in the end.

2. also the solute itself has a certain volume inside the box.

Since I mainly use AMBER, I use the leap module within AmberTools to add ions. But as it was said, you can also use VMD, etc.

For sure, if your protein is neutral and you are supposed to have let's say 50 Ca+2, you should add also 100 Cl- for neutralization. in fact, where these Ca2+ ions come from in the real system? What other neatively charged ions are supposed to be present in your solution?

Otherwise, if you don't want any negatively charged ion at all in your system an alternative could be (in AMBER, I don't know in GROMACS) to use a uniform neutralizing plasma - i.e., since when you use PME the system should be neutral, if your simulation system has an overall charge, this residual charge is automatically spread over all the atoms in the system. Depending on how large your system is, this can result in a very small perturbation. See, for example, this AMBER forum for details:

http://archive.ambermd.org/201112/0544.html

but I believe it should be the same also in GROMACS! See:

http://gromacs.5086.n6.nabble.com/uniform-neutralizing-plasma-for-PME-td4427622.html

from MD point of view its bette to add counter ions, as it will maintain the system

neutral and this will be good for using PME method to evaluate electrostatics.

on the other hand you need to look in the literature whether this calcium concentration is realistic or not....to introduce ions in the system you can use VMD .......adding ions interface.....hope this works....

check this link for adding ions. http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin/gmx-tutorials/lysozyme/04_ions.html

and you have to neutralize the system. If you do not want cl ions then add some anionic lipids to match the charge of the calcium.

Physiological salt concentrations are around 100mM, so I 400nM calcium does not strike me as high at all. How many Ca2+ ions would you actually have in a (relatively small) simulation box with e.g. 200K water molecules? Zero on the average. Yes you should neutralize your system. What counterions does the physical system you are modeing have ?

I agree that 0.4 mM is very low. So if you calculate the volume of your simualtion box and calculate the number of ions you should introduce probbably (depending on the size of your system) you'll find out a very small number.

In principle, however, there are 2 points that deserve comments:

1. the initial size of your box will likely change during the MD, since yousually you may want to run an NPT simulation. So the Ca2+ conc. should be calculated with respect to the "final equilibrated" box, otherwise you'll have a higher Ca2+ in the end.

2. also the solute itself has a certain volume inside the box.

Since I mainly use AMBER, I use the leap module within AmberTools to add ions. But as it was said, you can also use VMD, etc.

For sure, if your protein is neutral and you are supposed to have let's say 50 Ca+2, you should add also 100 Cl- for neutralization. in fact, where these Ca2+ ions come from in the real system? What other neatively charged ions are supposed to be present in your solution?

Otherwise, if you don't want any negatively charged ion at all in your system an alternative could be (in AMBER, I don't know in GROMACS) to use a uniform neutralizing plasma - i.e., since when you use PME the system should be neutral, if your simulation system has an overall charge, this residual charge is automatically spread over all the atoms in the system. Depending on how large your system is, this can result in a very small perturbation. See, for example, this AMBER forum for details:

http://archive.ambermd.org/201112/0544.html

but I believe it should be the same also in GROMACS! See:

http://gromacs.5086.n6.nabble.com/uniform-neutralizing-plasma-for-PME-td4427622.html

@Ramakrishnan Natesan · Dear Ram, lipid idea is not bad, but i have a doubt , during simulation lipid molecules may hinder intramolecular interactions of AA in peptide. Kindly clear my doubt if is it really a genuine.

@Giovanni Pavan, The calcium ion enter via a ATP-mediated pump, So I dont know if there are counter ion going in or out to neutralise in actual cellular system.

How do you suggest calculating calcium ion conc in final equilibrated box ? and make necessary changes ?

Thanks for interesting answers

400nM (0.0004mM) Ca2+ means one calcium ion in about 140 million water molecules, which you probably do not want to simulate (water is about 55M). If you want to do all atom MD, you will most likely have one Ca2+ in the box (with an effective concentration that is much higher).

ohh ... I am extremely sorry, The calcium ion conc in 400 mM (milli molar).

if the concentration is 400 mM, what I do usually is to calculate the box size volume.

Usually in the first steps of MD in NPT conditions the box shrinks. In my experience, in the case of medium size proteins solvated in octahedric TIP3P box with 14 ang. buffer the volume shrink is roughly 10% - so I would simply calculate the number of necessary ions and I would put just the 90% of the total number.

However, of course you should take this indication as just indicative, as these numbers can change depending on the solute (for highly charged dendrimers the shrink is larger).

Thus, is you need high percision on the Ca+2 concentration I can:

1. run an NPT MD simulation of your protein only in water

2. check the final equilibrated box volume (typically this takes a few nanoseconds)

3. calculate the number of Ca+2 based on this volume value

4. insert the ions and start from the beginning

however, also what suggested by Leandro - to calculate Ca+2 number based on the number of water is sensible.