the results obtained after docking using RosettaDock and ZDOCK

Hello,

When performing unbound protein docking using RosettaDock, I first conduct global docking with ZDOCK, selecting the top 10 results, and subsequently subject each of them to docking using RosettaDock. In the RosettaDock protocol, I perform prepack using the prepack protocol for structure preprocessing, followed by local docking using the docking_protocol to generate 1000 conformations. Subsequently, I sort the structures based on the total_score, selecting the lowest I_sc scores from the top 5%, refine the top 10 structures locally. The refined structures are then scored using the zrank2 scoring function, and the one with the lowest score is chosen as the final prediction.

尽管遵循此协议，但与输入结构相比，我观察到 Irmsd 和 Lrmsd 的改进很小。

我正在使用以下标志：

$ROSETTA_BIN/docking_prepack_protocol.linuxgccrelease @flag_prepack

-s 1FLE_z1.pdb

-合作伙伴E_I

-dock_ppk

-ex1

-ex2aro

-unboundrot 1FLE_z1.pdb

$ROSETTA_BIN/docking_protocol.linuxgccrelease @flag_local

-in：file：s 1FLE_prepack.pdb

-unboundrot 1FLE_prepack.pdb

-nstruct  1000

-合作伙伴E_I

-dock_pert 3 8

-ex1

-ex2aro

-out：路径：全部 ./local_out/

-out：后缀_local

$ROSETTA_BIN/docking_protocol.linuxgccrelease @flag_refine

-in：文件：s local_save/1FLE_prepack_local_0001.pdb

-nstruct  20

-合作伙伴E_I

-docking_local_refine

-use_input_sc

-ex1

-ex2aro

-out：文件：fullatom

-out：路径：全部 ./refine_out/

-out：后缀 _refine

如何使用 RosettaDock 从 ZDOCK 细化预测的构象，以及如何使用 total_score 和 I_sc 来选择构象？

谢谢。