Examples of binding site predictions from CASP8 targets using FunFOLD. The green sticks represent residues in the model that FunFOLD has predicted as binding to the ligands. The red sticks represent residues that were not predicted or incorrectly predicted as potential ligand binding residues. The blue sticks represent the observed ligand binding site residues in the experimental structure. The white spheres and the thin white sticks represent ligands either predicted (A and C) or observed (B and D). A) An example of a good FunFOLD prediction for CASP8 target T0407 (3e38), with the predicted binding site residues and ligands shown. The FunFOLD method predicted both metals and SO4 in the ligand cluster (CL-3, ZN-6, SO4-2 and FE-2), with the centroid ligand predicted to be SO4. B): The predicted binding site for T0407 using the model superposed onto the experimental structure, with both the observed and predicted binding site residues shown. The observed binding site ligands are also shown (ZN - 3). The red sticks represent the under-prediction of HIS157. C) An example where FunFOLD under-predicts to a greater extent for CASP8 target T0483 (3dls). The predicted ligand binding site residues and ligands are shown. The FunFOLD method predicted the centroid ligand as a nucleotide (ANP), with the ligand cluster containing both nucleotides and metals (ANP-3, ATP-1, STU-1, ADP-1, MG-5 and MN-2). D) The predicted binding site of the top model for T0483 superposed onto the experimental structure, with both the observed and predicted binding site residues shown. The binding site ligands ADP and Mg-2 are also shown.