Hi Nada,
As Oliver pointed out, the plots using the pink and green colors are the results of the calculated Gaussian curvature. Positive Gaussian curvature values are coloured in green, and negative values are coloured in pink. Negative Gaussian curvature represents saddle points, positive Gaussian curvature represents concave regions of the membrane. The plots obtained with MembraneCurvature do not provide information about lipid densities, only about surface, mean and Gaussian curvature.
Regarding your second question, I'll try to clarify using the tutorial you used as a reference.
The example provided uses a POPC:POPE:CHOL membrane using files of an MD simulation that was performed using the Martini force field. Since we use a coarse grained description for the system, we take our atoms of reference as the phosphate groups of all the lipids in the simulation box, i.e. PO4 beads. Then we do:
PO4_headgroups = universe.select_atoms('name PO4') # selects all lipid types that has a PO4 bead,
Please note that the selection above does not ask for specific lipids.
If you would like to see the surface formed POPC only, then the selection should look like:
PO4_POPC = universe.select_atoms('resname POPC and name PO4') # selects the PO4 beads of POPC molecules only
Then if you run membrane curvature for the selection above, you can find the associated surface formed by POPC. Please be aware that with this approach your sampling may be reduced and therefore populating the grid to derive surfaces and membrane may not be accurate (or possible at all) if the number of reference atoms is limited.
As for the scale in the plot, the examples provided in the tutorials are plotted with matplotlib. For more information on how to customize the colorbar, please visit matplotlib docs where you can find several examples:
https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot.colorbar.html
Hope this helps.