Back to Mover page.
King C, Garza EN, Mazor R, Linehan JL, Pastan I, Pepper M, Baker D. Removing T-cell epitopes with computational protein design. Proceedings of the National Academy of Sciences. 2014 Jun 10;111(23):8577-82. http://www.pnas.org/content/111/23/8577.long
Nivón, L. G., Bjelic, S., King, C. and Baker, D. (2014), Automating human intuition for protein design. Proteins, 82: 858–866. doi:10.1002/prot.24463 http://onlinelibrary.wiley.com/doi/10.1002/prot.24463/full
This mover will first attempt isolated/independent mutations defined in the input task operation, score/filter them all, rank them by score, then attempt to combine them, starting with the best scoring single mutation, accepting the mutation only if the filter score decreases (see skip_best_check for optional exception), and working down the list to the end. Optionally test one of the top N mutations at each positions instead of just the best.
This mover is parallelizable with MPI. To use it, you must set the option parallel=1, and you must set the command line flag nstruct = nprocs - 1
Note: Each attempted mutation is always followed by repacking all residues in an 8 Å shell around the mutation site. The user-defined relax_mover is called after that.
Note: Producing the very first output structure requires calculating all point mutant filter scores, which may take a bit, but output of subsequent structures (with nstruct > 1 ) will re-use this table if it's still valid, making subsequent design calculations significantly faster. However, the table must be recalculated each time if it is receiving different structures at each iteration (e.g. if movers that stochastically change the structure are being used before this mover is called).
<GreedyOptMutationMover name="(&string)" task_operations="(&string comma-separated taskoperations)" filter="(&string)" scorefxn="(score12 &string)" relax_mover="(&string)" sample_type="(low &string)" diversify_lvl="(1 &int)" dump_pdb="(0 &bool)" dump_table="(0 &bool)" rtmin="(0 &bool)" stopping_condition="('' &string)" stop_before_condition="(0 &bool)" skip_best_check="(0 &bool)" reset_delta_filters="(&string comma-separated deltafilters)" design_shell="(-1, real)" repack_shell="(8.0, &real)"/> #Pareto mode example, minimize filter 1 and maximize filter2 <GreedyOptMutationMover name="gopt" task_operations="task" relax_mover="min" scorefxn="score12"> <Filters> <AND filter_name="filter1" sample_type="low"/> <AND filter_name="filter2" sample_type="high"/> </Filters> </GreedyOptMutationMover>