All parameters for the three PSEUDO stages. CLI flags and YAML keys share the same names.
Debias (pseudo-debias)
debias — core omission parameters
Parameter
Type
Default
Description
run_name
str
required
Unique identifier for the run. Used as the top-level output directory name.
omit_type
str
"atoms"
Structural unit to omit per iteration. Options: amino_acids, atoms.
omit_fraction
float
0.1
Fraction of the structure omitted per iteration. Range: (0, 1).
iterations
int
5
Number of stochastic omission iterations.
always_omit
str
null
Comma-separated selection string for atoms/residues omitted in every iteration, e.g. "A 567, A 234".
seed
int
42
Random seed for reproducibility of the omission mask.
structure_path
str
null
Absolute path to the input PDB or CIF file. Required if screening_path is not set.
reflections_path
str
null
Absolute path to the input MTZ reflections file. Required if screening_path is not set.
screening_path
str
null
Path to a CSV or SQLite (Diamond SoakDB) screening file for batch processing.
sqlite_outcomes
str
null
Comma-separated substrings matched against RefinementOutcome in SoakDB files. No effect on CSV input.
max_structures
int
null
Cap on structures taken from a SQLite file. No effect on CSV or single-structure input.
screening_chunk_size
int
1000
Maximum number of omission jobs per sbatch array submission. Omission manifests are split into chunks of this size and submitted sequentially to avoid flooding the scheduler.
mtz_f_labels
str
null
Override auto-detected observed-data labels. Comma-separated amplitude+sigma pair, e.g. "FP,SIGFP" or "IMEAN,SIGIMEAN". Set when auto-detection fails or picks the wrong array.
mtz_rfree_label
str
null
Override auto-detected R-free flag column, e.g. "FreeR_flag". Set when the MTZ contains multiple flag columns or detection fails.
simulated_annealing
bool
true
Enable simulated annealing during phenix.composite_omit_map refinement. Disable for faster, lower-quality runs or when annealing causes instability.
Either structure_path + reflections_pathorscreening_path must be provided.
slurm — cluster resources
Parameter
Type
Default
Description
job_name
str
"debias_job"
SLURM job name shown in the queue.
partition
str
"cs05r"
Cluster partition to submit to.
time
str
"3-00:00:00"
Maximum wall time (D-HH:MM:SS).
mem_per_cpu
str
"5G"
Memory per CPU core.
cpus_per_task
int
1
CPU cores per task.
num_nodes
int
1
Number of compute nodes.
phenix_version
str
"1.21.2"
Phenix module version loaded in the SLURM environment (module load phenix/<version>).
paths — filesystem
Parameter
Type
Default
Description
work_dir
str
CWD
Root directory for all output.
Quantify (pseudo-quantify)
Configured entirely via CLI flags (no YAML):
Flag
Short
Type
Default
Description
--input_path
-p
path
required
Workspace directory produced by pseudo-debias.
--stem
-s
str
auto
Explicit experiment stem. Inferred automatically if omitted.
--k_factor
-k
float
1.0
Radius multiplier K for atom ownership spheres.
--map_cap
-c
int
50
Limit processing to the first N maps (for convergence testing).
--force
-f
flag
False
Overwrite existing results.
Analyse (pseudo-analyse)
Configured entirely via CLI flags (no YAML):
Flag
Short
Type
Default
Description
--input_path
-p
path
required
Workspace directory (single experiment or screening root).
--stem
-s
str
auto
Explicit experiment stem.
--map_path
-m
path
auto
Override the auto-discovered SNR map with a custom CCP4 file.
--model_path
path
auto
Override the processed model with a custom PDB/CIF file.
--k_factor
-k
float
1.0
K factor used to locate the SNR map in quantify_results/.
--map_cap
-c
int
50
Map cap used to locate the SNR map. Pass 0 to auto-detect.
--num_processes
-n
int
1
Parallel worker processes for screening mode.
--significance_alpha
-a
float
0.05
Significance level α for the null-distribution SNR threshold.
Screen Report (pseudo-screen-report)
Regenerates the HTML screen report from existing analyse_results/ data without re-running analysis. Configured entirely via CLI flags:
Flag
Short
Type
Default
Description
--input_path
-p
path
required
Root directory of the screening run.
--open_browser
flag
False
Open index.html in the default browser after generation.
Outputs written to <input_path>/:
File
Description
index.html
Interactive HTML report sorted by OPIA descending
metadata/{stem}_screen_result.json
Full per-crystal result dictionary
metadata/screen_summary_{timestamp}.json
Run-level summary (experiment list, mean OPIA, counts)
coot/{stem}_coot_view.py
Coot session script loading scored model + all colourcoded maps
MUSE scoring parameters (MUSEConfig)
These are Python-only parameters, passed to run_muse() via a MUSEConfig object. They control the internal MUSE scoring algorithm used by pseudo-analyse.
DensityScoreConfig
Parameter
Type
Default
Description
zeta
float
1.2
Upper truncation cap for normalised density values.
use_truncation
bool
True
Apply the upper cap. Set False to pass raw values.
MapNormalizationConfig
Parameter
Type
Default
Description
normalize
bool
False
Apply (ρ − μ) / σ normalisation. Set True for standard 2Fo-Fc maps.
global_mean_override
float|None
None
Override the computed map mean.
global_sigma_override
float|None
None
Override the computed map sigma.
AggregationConfig
Parameter
Type
Default
Description
ediam_shift
float
0.1
Shift s in the MUSEm power mean.
ediam_exponent
float
−2.0
Exponent p in the MUSEm power mean (negative = soft-minimum).
opia_threshold
float
0.8
MUSE score threshold for OPIA counting.
clash_threshold
float
0.1
Fractional sphere overlap that triggers a clash flag.
unaccounted_density_threshold
float
0.2
|MUSE−| above which unaccounted density is flagged.
missing_density_threshold
float
0.8
MUSE+ below which missing density is flagged.
GridConfig
Parameter
Type
Default
Description
max_spacing_angstrom
float
0.7
Maximum grid spacing (Å); finer grids are oversampled.