IRMSD Filtering Large-Scale Proteins

[kiamehrk]

Hello,

I have two large Vaccinia proteins (p4a and p4b) that I am trying to dock using a multi-stage (global and local) docking approach with RosettaScripts.

My concerns are:

1. I would like to set an IRMSD filter that is neither too strict to the point that it never passes nor too lenient so that I get too many poor quality docks at the end of the run.
2. Is there a way to filter by I_sc as well? I couldn't find an I_sc filter in Rosetta's documentation.

I currently have an IRMSD threshold of 20 at the global docking stage and 15 at the local docking stage. I would like to know the best method to either discover the best IRMSD threshold for my proteins or a standard threshold that is most commonly used, considering that the proteins I have are rather large.

Here are my scripts:

Global Docking:

<FILTERS>
    <IRmsd name="irmsd_filter_global" jump="1" threshold="20" scorefxn="fullatom_score"/>
</FILTERS>

<MOVERS>
    <DockingProtocol name="global_dock" low_res_protocol_only="1" partners="AB_CDEFGH"/>
    <SwitchResidueTypeSetMover name="switch" set="fa_standard"/>
</MOVERS>

<PROTOCOLS>
    <Add mover="global_dock"/>
    <Add mover="switch"/>
    <Add filter_name="irmsd_filter_global"/>
</PROTOCOLS>

Local Docking:

<FILTERS>
    <IRmsd name="irmsd_filter_local" jump="1" threshold="15" scorefxn="ref2015" />
</FILTERS>

<MOVERS>
    <DockingProtocol name="local_dock" docking_local_refine="1" dock_min="1" partners="AB_CDEFGH"/>
    <FastRelax name="FastRelax" scorefxn="ref2015" repeats="5"/>
    <InterfaceAnalyzerMover name="interface_analyzer" scorefxn="ref2015" pack_separated="0"/>
</MOVERS>

<PROTOCOLS>
    <Add mover="local_dock" filter_name="irmsd_filter_local"/>
    <Add mover="FastRelax" filter_name="irmsd_filter_local"/>
    <Add mover="interface_analyzer"/>
</PROTOCOLS>

[ajasja]

I would guess this is very system specific. I would run a small batch with higher RMSD, look through a few of the docks and get a feeling for a cutoff that way.