~This page contains descriptions of the most commonly-used pandda options (or giant toolkit scripts) ~
Don't know what a command-line option (or a whole script) does? -- see below
Running most programs with --show-defaults will give you the different options that can be used with the program.
Running with -h will list descriptions of the command line options. Running with -hh or -hhh will list more and more information about the parameter, such as what type of values may be provided for that parameters (a number, a string of letters, etc...).
Running with -a will list all possible options (including "expert" and "developer" options -- these are not normally needed).
For options with a "choice", the default options are marked with a "*", e.g. variable = *this_one_is_default this_one_is_not.
For some variables, multiple choices can be given, using a "+" sign to join, e.g. variable=a+b+c.
All command line arguments are called with flag=option. The most used options are discussed in the individual sections below.
Description:
pandda.analyse analyses a series of crystallographic datasets. This analysis reveals
Options:
| Flag | Type | Default | Function | Version |
|---|---|---|---|---|
| exclude_from_characterisation | Comma-separated list of dataset labels | None | Exclude datasets from calculation of the mean map, and other statistical maps. This is normally used when the dataset is known to contain a bound ligand. | v0.2+ |
| exclude_from_z_map_analysis | Comma-separated list of dataset labels | None | Exclude datasets from z-map analysis. This is normally used when the dataset is known not to contain a bound ligand. | v0.2+ |
| existing_datasets | choice: reprocess reload ignore | reload | Controls whether datasets that have been previously added are reloaded/reprocessed by the program. If datasets are not reloaded, the output hit-list will not include hits from these previous datasets. | v0.2+ |
| recalculate_statistical_maps | choice: yes no extend | extend | If a pandda has previously been run, the statistical maps will be reused to save time. This flag controls whether they are recalculated. This flag should normally be used with the reprocess_existing_datasets=True flag. |
v0.2+ |
| min_build_datasets | Integer | 40 | Minimum number of datasets to average to generate the mean map of the ground state crystal. Should normally be greater than ~30-40. | v0.1+ |
| max_build_datasets | Integer | 60 | Maximum number of datasets to average to generate the mean map of the ground state crystal. | v0.1+ |
| dynamic_res_limits | True or False | True | Allows the program to automatically determine the maximum and minimum resolutions to be analysed. | v0.1+ |
| high_res_upper_limit | A Number | 0.0 | Highest resolution to analyse | v0.1+ |
| high_res_lower_limit | A Number | 4.0 | Lowest resolution to analyse | v0.1+ |
| high_res_increment | A Number | 0.05 | Shell width when moving from high_res_upper_limit to high_res_lower_limit |
v0.1+ |
Description:
pandda.export takes the output from pandda.analyse and prepares it for crystallographic refinement.
Options:
| Flag | Options | Function |
|---|---|---|
| pandda_dir | File Path | output directory from pandda.analyse and pandda.inspect |
| export_dir | File Path | output directory for pandda.export |
| select_datasets | Comma-separated list of dataset labels | If you only wish to export a sub-selection of datasets, you can define them with this option. Dataset names must match the names of folders in the pandda directory. |
| generate_restraints | True or False (default True) | If True, then giant.make_restraints will be run on each of the generated ensemble structures. |
Bundled with panddas are a set of scripts under the name of "giant.script_name". These are separate from panddas, but can be useful for processing the output.
These scripts are largely intended to enable the refinement of crystallographic ensembles: generating, restraining, refining and splitting multi-state models.
| Script | Function | Description |
|---|---|---|
giant.datasets.cluster |
Generate groups of similar PDB or MTZ files | Clusters PDB or MTZ files by spacegroup and unit cell. Output can be directed to an output folder, where datasets will be organised by cluster. Dendrograms of the clustering are generated for each spacegroup, and then for each cluster. This is useful for organising the input to pandda.analyse. |
giant.make_restraints |
Generates parameter restraints for refinement | Automatically generates groups of residues whose occupancies should be the same during refinement. Residues with the same conformer IDs are grouped by proximity, and their occupancies are restrained. This script is designed to work on the output from giant.merge_conformations (and from pandda.export). |
giant.quick_refine |
Wrapper script for starting crystallographic refinements | Start refinements with REFMAC or phenix, and automatically organise the output. Takes the restraints output by giant.make_restraints. |
giant.score_model |
Calculates quality statistics for ligands | Uses EDSTATS to score ligand residues against the electron density. Also calculates B-factors of the ligand and surrounding residues for comparison. Can be run on multiple datasets simultaneously using giant.score_model_multiple or giant.score_model_set. |
giant.split_conformations |
Split an ensemble model in its different states | For a multi-conformer model of a protein, this will output several models for the different conformer IDs. Structures can be split by conformer ID, or by presence of a ligand (states with a ligand and states without a ligand). |
giant.merge_conformations |
Merge two models representing different crystal states (e.g. bound v unbound). | Generate an ensemble model from two discrete crystal states. Single-conformer parts of the structure that are the same between models will result in one conformer. Residues which move or are only present in one structure will be given appropriate conformer IDs. |
giant.local_align |
Align structures to a reference (using the alignment method in pandda.analyse). | Uses a local alignment algorithm to align two conformations of the same structure. Not guaranteed to maintain reasonable geometry (e.g. bond lengths). |