Author(s)

• anomalous_signal: Tom Terwilliger

Purpose

anomalous_signal is a tool for estimating the anomalous signal in a SAD experiment and for predicting whether this signal is sufficient to solve the structure. This tool normally is used in combination with plan_sad_experiment (planning the experiment) and scale_and_merge (scaling the unmerged data and also obtaining two half-datasets).

Usage

How anomalous_signal works:

• You supply anomalous_signal with scaled anomalous data, two half-dataset files with scaled anomalous data, and the number of sites or a sequence file and name of the anomalously-scattering atom.
• anomalous_signal will calculate the anomalous signal in your dataset from (1) the half-dataset anomalous correlation, (2) the skew of the anomalous difference Patterson map, and (3) the estimated measurement error in your data.
• anomalous_signal will then estimate the probability that you can solve this dataset using likelihood-based HySS (standard run) and will estimate the figure of merit of phasing that you should obtain.
• The way that plan_sad_experiment and anomalous_signal estimate the probability that you can solve this dataset is to compare the anomalous signal in this dataset with the anomalous signal in other datasets at the same resolution. Then the fraction of similar datasets that can be solved by HySS is used as the probability that your dataset will also be solved.

Output from anomalous_signal

anomalous_signal provides a summary of the skew, measurement error, and half-dataset correlations in your dataset and the estimated anomalous signal, probability of solving the structure and estimated figure of merit of phasing. Here is an example:

-------------------Summary of signal in this dataset ------------------------


       Shell

                       CCano   Nrefl Nrefl
Resolution Esqr I/sigI  half   anom   half
48.2- 6.0  0.09  33.09  0.85    2090  2032
 6.0- 5.5  0.34  23.31  0.53     667   659
 5.5- 5.0  0.34  22.84  0.49     956   947
 5.0- 4.5  0.28  29.23  0.38    1468  1447
 4.5- 4.0  0.42  24.00  0.16    2305  2276
 4.0- 3.5  0.57  14.02  0.05    3832  3751
 3.5- 3.0  0.84   5.91  0.01    6869  6494
 3.0- 2.9  1.52   2.39  0.00    2035  1734

       Cumulative


----------------------Data quality-----------------    Best guess of expected
                                                      results of finding sites
                                                     ------ and phasing--------

                     CCano   Nrefl                    P(Substr)
Resolution Skew Esqr  half   anom    CC* Signal  +/-     (%)       FOM*  +/-
48.2- 6.0  0.03 0.08  0.85    2090  0.71   6.3   0.4      23       0.3   0.1
48.2- 5.5  0.02 0.10  0.82    2757  0.76   7.8   0.3      40       0.3   0.1
48.2- 5.0  0.01 0.12  0.78    3713  0.70   8.3   0.6      47       0.3   0.1
48.2- 4.5  0.03 0.14  0.72    5181  0.68   9.5   0.7      61       0.3   0.1
48.2- 4.0  0.03 0.17  0.62    7486  0.72  12.1   0.6      76       0.4   0.0
48.2- 3.5  0.02 0.25  0.43   11318  0.66  13.6   1.4      83       0.4   0.0
48.2- 3.0  0.00 0.47  0.24   18187  0.48  12.7   2.0      79       0.3   0.1
48.2- 2.9  0.00 0.59  0.22   20222  0.42  11.8   1.7      74       0.3   0.1

Notes:
Skew is skew of origin-removed anomalous difference Patterson
Esqr is the squared ratio of sigmas to differences:<SigAno**2>/<Dano**2>
I/sigI is the mean value of (Ihkl/sigIhkl)
CCano half is the correlation between half-dataset anomalous differences
Nrefl half is the number of anomalous differences in both half-datasets
Nrefl anom is the number of anomalous differences in the entire dataset
CC* is estimate of anomalous correlation to ideal data
Signal +/- sigma is estimate of anomalous signal (peak height at coordinates of
anomalously-scattering atoms in difference Fourier phased with model
phases.)

Note that the anomalous signal increases with the square root of the
number of reflections and decreases with the square root of the number
of sites.

P(Substr) is estimate of probability that the sub-structure can be found
with LLG-based HySS and this data.
FOM* +/- sigma is estimate of phasing FOM if sub-structure is solved

Possible Problems

anomalous_signal assumes that your crystal is similar to others that have been solved. If your crystal has serious decay or other factors that reduce signal to noise, the estimates provided by plan_sad_experiment may be too optimistic.

Literature

List of all available keywords

• input_files
  • data = None Anomalous data (I+ and I- or F+ and F-). You must supply unmerged or merged data plus two half-dataset datafiles (usually these half-dataset files are obtained with phenix.scale_and_merge) Any standard format is fine.
  • data_labels = None Optional label specifying which columns of anomalous data to use. Not necessary if your input file has only one set of anomalous data. .short_caption = Data labels
  • half_dataset_a = None Data for half-dataset A. Normally required if the main dataset is merged. This half-dataset normally comes from running phenix.scale_and_merge on unmerged data to produce half_datasets A and B. Any standard format is fine.
  • half_dataset_a_labels = None Optional label specifying which columns of anomalous data to use. Not necessary if your input file has only one set of anomalous data. .short_caption = Half-dataset A labels
  • half_dataset_b = None Data for half-dataset B. Normally required if the main dataset is merged. This half-dataset normally comes from running phenix.scale_and_merge on unmerged data to produce half_datasets A and B. Any standard format is fine.
  • half_dataset_b_labels = None Optional label specifying which columns of anomalous data to use. Not necessary if your input file has only one set of anomalous data. .short_caption = Half-dataset B labels
• output_files
  • scaling_log = scale.log Output log file for scaling (if applied)
  • output_file = scaled_data.mtz Output data file name for merged scaled data. Output files are produced if half-dataset files are not supplied and unmerged input data is scaled and divided into half datasets half-dataset files are not supplied
  • output_half_dataset_a = half_dataset_a.mtz Output data file name for half dataset A. Output files are produced if half-dataset files are not supplied and unmerged input data is scaled and divided into half datasets half-dataset files are not supplied
  • output_half_dataset_b = half_dataset_b.mtz Output data file name for half dataset B. Output files are produced if half-dataset files are not supplied and unmerged input data is scaled and divided into half datasets half-dataset files are not supplied
• directories
  • temp_dir = None Optional temporary work directory. If specified, it will be used for temporary files. If clean_up=True and it did not exist before, it will be removed at the end of the run. If it did exist, it will be left.
  • output_dir = "" Output directory where files are to be written
  • gui_output_dir = None GUI use only - does not apply to command line version
• crystal_info
  • resolution = None high-resolution limit
  • low_resolution = None low-resolution limit
  • space_group = None Space group (normally read from the data file)
  • unit_cell = None Unit Cell (normally read from the data file)
  • seq_file = None Optional sequence file (1-letter code). Separate chains with a blank line or line starting with > If not supplied, you will need to supply the number of sites with sites=xxx.
  • chain_type = *PROTEIN RNA DNA Chain type (PROTEIN RNA DNA). This is used to estimate the number of atoms from the number of residues. If not supplied, you will need to supply the number of sites with sites=xxx.
  • atom_type = None Name of anomalously-scattering atom. If not supplied, you will need to supply the number of sites with sites=xxx.
  • sites = None The number of anomalously-scattering atoms in the asymmetric unit.
  • b_value_anomalous = None Estimated Wilson B-value for the anomalously-scattering atoms. Normally leave as None and it will be estimated from Wilson B
• analysis
  • half_dataset_cc = True Analyze half-dataset anomalous correlation. Requires unmerged data file or half_dataset_a and half_dataset_b files, each with half of data from an umerged dataset. The half-dataset files normally come from running phenix.scale_and_merge on unmerged data to produce half_datasets A and B.
  • skip_scaling = False Skip scaling and use merging_statistics to get half-dataset CC.
  • resolution_dependent_predictors = True Use Bayesian predictors that are resolution-dependent
  • planned_number_of_datasets = None You can predict the useful anomalous correlation, anomalous signal, and figure of merit of phasing that you would expect if you collected a total of N datasets like this one. See also ideal_cc_anom which is the estimated maximum anomalous correlation achievable with these crystals.
  • ideal_cc_anom = 0.75 Estimate of useful anomalous correlation for perfectly measured data. Used as maximum anomalous correlation achievable when planned_number_of_datasets is set.
• control
  • verbose = False Verbose output
  • clean_up = True Clean up and remove temp directory after finishing
  • job_title = None Job title in PHENIX GUI, not used on command line