Comparison of aggregation predictions and experimental available data for human globular proteins and proximity of aggregation-prone regions to predicted and real interfaces.

1Percentage of residues in the aggregation-prone region at less than 3 Å from a protein predicted interaction residue.2Percentage of residues in the aggregation-prone region at less than 3 Å from a residue located at the interface of the following complexes: β2-microglubulin in complex with HLA heavy chain [1DUZ] and with HFE [1A6Z]. Native tetrameric structure of transthyretin (PDB code 1TTA). Dimeric structure of SOD1 (PDB code 2C9V). Lysozyme in complex with a camelid antibody (PDB code 1OP9). Dimeric structure of Immunoglobulin LC variable domain (PDB code 2Q20). HCs and LCs of a IgG1 human immunoglobulin (PDB code 1HZH).1,2In brackets the percentage of residues in the aggregation-prone region close to a random surface of the same size than the considered interface.