I am trying to insert cholesterol inside the DPPC membrane by using gmx insert_molecules -replace. But number of DPPC molecules replaced is much higher than the number of cholesterol inserted? Please suggest me a way to build the system by replacing desired number of DPPC.
You can play around with gmx insert-molecules flags for example by scaling (-scale) the vdw radii for insertion or having larger number of tries (-try). Take a look at the manual page (https://manual.gromacs.org/current/onlinehelp/gmx-insert-molecules.html).
My feeling however is that you wont get rid of the problem. You are trying to insert a bulky molecule such as cholesterol in an equilibrated DPPC membrane. This means that, even if the average area per lipid for a DPPC molecule may not be very far from that of a mixed chol/DPPC membrane, every DPPC lipid molecule alone is going to be disordered tails and you will have to remove more than one for having enough volume to put a cholesterol molecule, especially if you are in a liquid-like regime (temperature). And then you will have to re-equilibrate the system. Also, can you respect the symmetry of the bilayer, that is, insert the cholesterol molecules with the polar head group at the right depth? You will also have to try to substitute roughly the same number of lipid molecules per side as otherwise you will lose the leaflets symmetry.
Why do you want to substitute the molecules? My advice would be to start with a new membrane simulation where your content of cholesterol is already correct.
maybe you can start with this one:
https://ftp.gromacs.org/contrib/topologies/1-1_CHOL-DPPC.zip