You are here

Protein Interface Design using PSSM

2 posts / 0 new
Last post
Protein Interface Design using PSSM
#1

Dear Rosetta community, 

I have been worked using the rosetta scripts over past of few months on protein interface design. The issue is to design the interface of a protein (chain B) into a complex using the PSSM of this protein to drive the mutations. For that, I requested the FavorSequenceProfile before the design (see the script bellow ). A designed sequence was obtained but I am not sure if PSSM was requested during the process, because no task_operations was defined in PackRotamersMover. I tried to use the SeqprofConsenses for this, but it did not work. 

All suggestions are welcome!

Best,

Carlos

***********

<ROSETTASCRIPTS>

<SCOREFXNS>

        <ScoreFunction name="ref2015_toPSSM" weights="ref2015">

                 <Reweight scoretype="res_type_constraint" weight="0.4"/>

        </ScoreFunction>

</SCOREFXNS>

<TASKOPERATIONS>

        <ProteinInterfaceDesign name="pido" interface_distance_cutoff="8.0"/>

</TASKOPERATIONS>

<MOVERS>

        <PackRotamersMover name="interface_design" scorefxn="ref2015_toPSSM" task_operations="pido"/>

        <FavorSequenceProfile name="FSP" pssm="%%pssm%%" scaling="none" chain="2"/>

</MOVERS>

<PROTOCOLS>

        <Add mover_name="FSP"/>

        <Add mover_name="interface_design"/>

</PROTOCOLS>

</ROSETTASCRIPTS>

***********

 

Category: 
Post Situation: 
Sun, 2020-05-24 14:27
ccruz

The SeqprofConsensus Task Operation should get you where you need to go. Just be sure to include it in the list of taskops you provide to the mover and include the res_type_constraint term in your score function, as you are.

<TASKOPERATIONS>
     <SeqprofConsensus name="pssm_cutoff" filename="%%pssm%%" min_aa_probability="0"  convert_scores_to_probabilities="0" probability_larger_than_current="0" keep_native="1"/>
</TASKOPERATIONS>
<MOVERS>
     <PackRotamersMover name="design" scorefxn="scorefunction"  task_operations="init,pssm_cutoff"/>
</MOVERS>

One possibility is that mutations in the interface never score better than the native residue. Historically, this is often the reason that a soft_rep score function is used in a first round of design, followed by the standard scoring function. Very commonly now, the FastDesign Mover is used because it ramps the repulsive term allowing for mutations to occur early in the protocol and then relax into acceptance later on. Check out the protocols at https://github.com/cdbahl/protein_design_protocols (their reference is this paper: Lau, Y.-T. K., Baytshtok, V., Howard, T. A., Fiala, B. M., Johnson, J. M., Carter, L. P., et al. (2018). Discovery and engineering of enhanced SUMO protease enzymes. The Journal of Biological Chemistry, 293(34), 13224–13233. http://doi.org/10.1074/jbc.RA118.004146)

Tue, 2020-05-26 06:33
nannemdp