dpocket-pocket volume reduces in the presence of HETATMs

[Peter-obi]

I am trying to calculate the volume of the binding pocket for a protein using dpocket in the presence and absence of lipids like POPC and POPE. The lipids are outside the binding site but within 4 angstroms of the ligand so I expected an increase in the pocket volume since I am using the explicit (-E) flag to consider all atoms within 4 angstroms of the ligand. However, the pocket volume is decreased in the presence of the lipids. For example, without the lipids, it is 2000 but in their presence it's 1600. I expected it to at least remain the same and not decrease except I am missing something here. I tried to'trick' the algorithm and changed the names to another name in the protected HETATM list to see if the problem was as a result of them not being considered by the algorithm but the results remained the same. How can I solve this?