FilterScan

Back to Filters page.

FilterScan

Described in Whitehead et al., Nat Biotechnol. 30:543

Scan all mutations allowed by task_operations and test against a filter. Produces a report on the filter's values for each mutation as well as a resfile that says which mutations are allowed. The filter can work with symmetry poses; simply use SetupForSymmetryMover and run. It will figure out that it needs to do symmetric packing for itself.

<FilterScan name="(&string)" scorefxn="(score12 &string)" task_operations="(comma separated list)" triage_filter="(true_filter &string)" dump_pdb="(0 &bool)" filter="(&string)" report_all="(0 &bool)" relax_mover="(null &string)" resfile_name="([PDB].resfile &string)" resfile_general_property="('nataa' &string)" delta="(0 &bool)" unbound="(0 &bool)" jump="(1 &int)" rtmin="(0&bool)" delta_filters="(comma delimited list of filters)" score_log_file="('' &string)"/>

triage_filter: If this filter evaluates to false, don't include the mutation in the resulting resfile
dump_pdb: dump models for each of the single mutants that pass triage filter. The pdb is scored and dumped after running the relax mover. The naming is: <input file name>RESxxxRES where the suffix for the Ala278->Arg mutant will look like ALA278ARG.
filter: Used only for reporting the value for the pose in the tracer report
report_all: By default, only attempted mutations which pass triage_filter will be evaluated by filter and reported in the tracer report. If report_all is true, report the value of filter for all evaluated mutations. (Note this will increase the number of calls to filter /the computational cost.)
relax_mover: After mutation, what mover to use for relax (minimization may be a good idea) This mover is in addition to repacking done as part of the mutation. (Repacking is done according to task_operations , but is limited to an 8 Angstrom shell around each mutated residue.)
scorefxn: The scorefunction to use with the mutation repacking
resfile_name: the output resfile name. Defaults to what's the -s on the commandline +".resfile"
resfile_general_property: What to do with all other residues in the resfile
delta: Test the filter against a baseline which is the filter's value at the start of the run?
unbound: Test the filter in the unbound state?
jump: If unbound, which jump?
rtmin: carry out an rtmin repack step before the relax steps? Improves the fit of the mutated residue but can lead to noisy energies.
delta_filters: a list of filters of type DeltaFilter. See below for baseline evaluation.
score_log_file: filename to put the score values (per residue filter output).

To compute a baseline, the entire pose is first repacked (with include current, no design, initialized from the comandline, but ignoring the taskoperations), and relax_mover is called. This can be useful to get the pose's energy down. If delta_filters are specified then before introducing mutations to any given position, the mutation to self is made (say Arg25->Arg), repacked/rtmin,relaxed as usual and each filter's value is evaluated. Then, these values are set as the baselines for each DeltaFilter.

Filter and triage_filter are potentially confusing. You can use the same filter for both. Triage_filter can be more involved, including compound filter statements, whereas the filter option is reserved to filters that have meaningful report_sm methods (ddG, energy...).

The reported values from filter will appear in a Tracer called ResidueScan, so -mute all -unmute ResidueScan will only output the necessary information

Use the unbound option only on a Prepacked structure with jump_number=1, o/w the reference energy (baseline) won't make any sense.

FilterScan

FilterScan

See also