I've stumbled upon an odd development while working the T-cell Lymphoma Invasion Protein.

The loop between 9-Isoleucine and 15-Phenylalanine is acting strangely. At the end of a rebuild script, the highest scoring structure had this loop literally bent back upon itself (see figure1.png). When I tried to rebuild it manually, it stretched out to freakish proportions, and the score dropped to negative infinity. (Well, close enough.) I was too flabbergasted to get a screenshot.

When I did manage to get the loop back into a more reasonable shape, I shook the sidechains. 11-Lysine decided to then bury itself INSIDE the backbone (see figure2.png and figure3.png for before and after photos).

I call "BUG!" This protein ain't behavin'.