In 2010, a large group at Abbott led by Philip Hadjuk reported a compound:
identified through screening that bound in one of our predicted druggable sites, confirmed by crystallography:
If you want to look for yourself, check out pdb code 3NEW. Though this may not have been the most difficult prediction to make, as mentioned in the original paper, another kinase had allosteric inhibitors bound near the same region, it is encouraging that the method can identify druggable sites, which also turn out to be allosteric as shown in the paper from Abbott.
Using the residues from the 1KV1 crystal structure that our druggability method predicted in surface and the aligned structure with ligand found by Abbott, here is a comparison showing the ligand really does bind in the pocket predicted:
