Construction of the N-Terminal Domain (NTD) Dimer

The NTD (residues 1–78) was modeled using the threading-based I-TASSER modeling server (27). I-TASSER identifies the best templates by employing LOMETS threading programs based on the Z-score. The PDB IDs used for modeling were 5T7Q_A, 3J9A_A, 2MK0_A, 3ZIF_R, 2V6L_Z, 3J65_Q, and 2FFT_A. The letter after the underscore is the chain identifier. Among the top five models, the model having the highest C-score (i.e., −4.11) was selected for protein–protein docking. The dimeric state of the NTD was predicted using ZDOCK with default parameters and without specifying any binding or blocking residues. The top ranking docked conformation was optimized by energy minimization.