Dear Bukunmi, extending simulation is fine using:

gmx mdrun -v -deffnm md -cpi md.cpt -append

But I don't understand why you want to increase the temperature. better do that separately by modifying the .mdp file. 

Dear Faez, thanks for your answer but i am studying the effect of temperature on a system so i need to warm it up to 600K but the resources are not available for one simulation, so i broke it down into two stages.. first to 500K and the next simulation should be to 600K.ordinary change in the mdp file treated the system based on its present situation at 500K but i need to preserve some system functions from 300K if you understand me.

You probably want to explore the simulated annealing options, to warm the system from 500 to 600 K. In this case, you need a new .mdp file with the relevant options and use the approach with grompp -t old.cpt to provide the previous simulation's state as the new starting point for the second run. You cannot simply extend the simulation because you are changing the simulation instructions and the properties of the ensemble.

thanks Justin, i am currently testing the simulated annealing option but i am confused about the addition of tpbconv

You won't use tpbconv (now called convert-tpr in more recent versions of the code). You're essentially just starting a new simulation from the starting conditions of an old one.

Hey all, let me ask a simple and stupid question. If the computer turns off before the simulation is over, and continue the simulation afterwards, how can I put the two trajetories together to analyse the final result in one simple trajetory?

Thanks once again Justin but are you saying that with the grompp -t old.cpt, i will be able to preserve the properties from the initial configuration at 300K and the new simulation will still be related to the initial configuration not forgetting that the cpt was from the completed simulation at 500K or the system is going to completely take on the behaviour at 500K?

Nothing about the 300 K system will be in the .cpt file from the end of the run that finished at 500 K. That .cpt file has all the state variables of the final frame of the first run. I'm not 100% clear on what you did in that first run, but the point is that whatever is in that .cpt file (coordinates, velocities, etc) is what becomes the new starting state for the new simulation.

Thank you very much Justin, there are some restraints on the system at 300K which should still be maintained even upto 600K so i believe i will have to run the simulation as just one instead of breaking it down.

Restraints are preserved as long as the .mdp file + topology specify that they should be applied. There should be no problem here regarding restraints.

Andre Silva Pimentel,

you can open 2 trajectories in VMD one after the other, and also you can use the append option to continue the simulation over writing all the topology files from the point where it stopped,