3.1 Definition of the PPI interface
The pEDNs of PD-1 and PD-L1 were obtained from the FMO calculation of the complex using the equation (2) or (3), from which the differential EDN was also obtained by simply subtracting them. In Figure 1-A, the surface with a zero value of this differential EDN is pictured (the pEDNs of PD-1 and PD-L1 on this surface are the same value). As mentioned above, this surface is naturally considered as the PPI interface. In Figure 1-B, the pEDN are visualized on this surface. We note that positions with a pEDN larger than 0.001 a.u. (colored region) are widespread on the PPI interface, indicating that the EDNs of PD-1 and PD-L1 widely overlap with each other. Positions where the pEDN is larger than 0.01 a.u. (blue or purple) mean to be a significantly large overlap of the EDNs caused by close contacts between amino acid residues of the two proteins. For example, position P1in Figure 1-B is caused by contact between K131 of PD-1 and Q66 of PD-L1 (distance: 1.74 Å). Similarly, position P2 is cause by contact between Y68 of PD-1 and D122 of PD-L1 (distance: 1.61 Å). At the other positions with large pEDN values, amino acid residues are also in close contact.
These results suggest that an undulating PPI interface involving numerous amino acid residues was reasonably determined according to their pEDNs. Additionally, by visualizing the value of the pEDN at the PPI interface, the range of atomic contacts between proteins was clearly represented.