Flexible Antibody (immunoglobulin G form)

This is an articulated antibody model (surface rendering of PDB:1HZH), where the antigen-binding domains (Fab) and effector domain (Fc) are connected by 5 interconnected rings (not disulfides and flexible loops) and features a hole for a 5mm neodymium magnet in each Fab roughly in the centre of the antigen binding region (approximately replacing the CDR3 loops of the two abutting variable regions).

NB The ring joining the two heavy chains broke off when I removed the supports. I am aware that there is a disulfide bond between them, so without said ring the model is inaccurate.

Needs work

This is a 'work in progress', but actually not: I made one, I don't need more, but there is room for improvement.
Specifically, the rings are 3 mm thick, which is too little.

I printed it on my filament printer, not my resin printer because

• IMO removing supports for protein is more annoying on resin prints than filament prints... every time I print a protein I consider upgrading to dual filament printing
• I wanted to use a tricolour filament as opposed to grey which would need painting... less fun than a miniature as I would have had to figure out where the heavy and light chains meet.

In other words, I was lazy, mkay? This had the issue that the rings on a plane perpendicular to the printing bed (filament printer) fail. So in my second printing attempt I printed at an 45° angle. Were I to do a 3rd design attempt I would thicken the rings and make them elliptical to avoid the rings breaking*.
A second change I would do is make the magnet holes shallower and more exposed. That there is a CDR2 loop that covers is cool, but makes the magnet's hold weaker —the model is not self-supporting even against pure iron strips (steel alloys are less ferromagnetic so not a chance).
A third change would be printing a bunch of extra rings for DIY fixes.

∗ I dropped the antibody and the Fab came off —macro-scale protein engineering.