Below is the final structure for the two antibodies bound to the target antigen starting with just a pair of antibodies and an antigen, as well as the predicted structure for NRP2 (front and back) respectively. And below them is the reference model with the antibodies in pink and the pentamer truncated in white.
The published structures of 7M30 and 7M22 were then superpositioned over the two predicted structures to calculate the RMSD values in Pymol. 7M30(rainbow by chain) with the predicted structure(white), giving an RMSD of 9.482
It is important to note that the RMSD of 9.482 is highly variable as the three dimensional coordinates do not align well because of antibody 7M1C and the flexible region. Therefore by truncating the unaligned portions of the homology model I am able to do a pseudo local refinement to get a better RMSD. When the predicted model for 7LYV (the top antibody) was truncated to match the pdb 7M30, the RMSD value lowered to 1.547 as shown below. To note, the blue structures that end away from the antibody exist in the flexible region of both the reference structure and the chosen starting structure. 7M30 is the structure with the dark blue flexible region.
A similar result is shown when the top portion of the predicted model is truncated to ensure only the pentamer backbone and the 7M1C antibody are present. The predicted model is in white while the reference structure is colored in a rainbow fashion by chain. The RMSD of this structure is 2.538 due to the antibody alignment.
Finally the superposition of 7M22(in white) and the predicted NRP2 structure(rainbow) showed an RMSD of 0.011 on the left and rotated 180 degrees on the right.