How do our predictions match up on real structures? Check out examples from our paper (red = correctly predicted kinks, blue = incorrectly predicted kinks, green = kinks not predicted). Legend found in original paper under supplementary figure S2.
Use our predictions? Cite our publication Meruelo AD, Samish I, and Bowie JU. TMKink: A method to predict transmembrane helix kinks. Protein Sci. 20:1256-64, 2011.