CTRUNCATE (CCP4: Supported Program)


ctruncate - Intensity to amplitude conversion and data statistics.


ctruncate -hklin filename -hklout filename -seqin sequence file name -colin column label -colano column label -colout column label -nres number of residues -no-aniso turn off anisotopy correction -amplitudes use amplitudes -comp if set to nucleic use DNA/RNA reference curve -prior force use of WILSON or FLAT prior -freein retain freeR flag
[Keyworded input]


Conversion of measured intensities to structure factors is complicated by the fact that background subtraction can result in negative intensities for weak reflections. CTRUNCATE is a new CCP4 program, intended ultimately to replace the original TRUNCATE program, which uses Bayesian statistics to calculate positive structure factors from negative input intensities. Small positive intensities are also boosted by the conversion process. The algorithm used is that of French and Wilson.

In addition, CTRUNCATE calculates a number of statistics from the intensity data, such as moments, cumulative intensity distributions and the Wilson plot. When output in graphical form, these can be used to assess data quality and to check for possible twinning.

CTRUNCATE looks for anisotropy in the data and performs anisotropy correction. A test for translational NCS is also performed. Two methods are used to test for twinning: Yeates' H test and Padilla and Yeates L test. The twinning tests, as well as the moments and cumulative intensity plots are done using data which has been corrected for anisotropy. However the output structure factors and the Wilson plot use uncorrected data. Should tNCS or twinning be found a flat prior based upon the same approximations as the Wilson prior is applied.


-hklin filename

(Compulsory) Input mtz file name. '-mtzin' is accepted as an alternative to '-hklin'.

-hklout filename

Output mtz file name. Default 'ctruncate_out.mtz'. '-mtzout' is accepted as an alternative to '-hklout'.

-seqin sequence file name

Input sequence file in any common format (e.g. pir, fasta), used for scaling.


-colin column label

Column names for mean intensity data. If only anomalous data is input, then only the anomalous data will be considered.

-colano column label

Column names for anomalous data. Both positive and negative names need to be specified e.g. '/*/*/[I(+),SIGI(+),I(-),SIGI(-)]'. If column names for anomalous data are not specified, the program will only consider mean data.

-colout column label

Identifier to be appended to output column names.

-nres number of residues

Specifies the number of residues in the asymmetric unit for scaling purposes. If not used and an input sequence file is not provided, the number of residues will be estimated assuming 50% solvent content.


Anisotropy correction will not be performed.


Input data is structure factors rather than intensities.

-comp nucleic

Use DNA/RNA reference curve. The reference curve derived from 65 resonably high resolution intensities and structure factors from the PDB and nucleic acid data bank.

-prior WILSON or FLAT

Force use of WILSON or FLAT prior.

-freein colname

Retain the the freeR column. Default colname is '/*/*/[FreeR_flag]'

Release Notes

- follow lead of K.Diederichs and extend tabulated range for truncate procedure to -37 (this is were R calculation becomes unstable for centric reflections)

- use range for anomalous limits rather than monatomical decrease

- protect output in cases were there are no twinops

- preserve MTZ history from input file

- obtain twinning fraction estimate from L-test (PY)

- reset anisotropy corrected data for moments

- reconfigure twinning tests
- add ML Britton

- fix memory corruption on ubuntu
- correct residue based cell contents correction

- use weighted mean and sig_mean=sqrt(s1*s1+s2+s2) for I and F
- not using combined sig_mean as this gives a large value for sigma on Fmean and Imean

- output missing FMEAN and SIGFMEAN when output DANO

- rationalise computation of means and sigmas from anomalous data
- use I/sigI derived data for F/sigI if it is given with anomalous flag

- correct anomalous measurebility plot and phil plot for anonalous input

- allow use of friedal pairs, without Imean

- use mean anisotropic eigenvalue when computing poorest resolution cuttoff

- fix issue of going out of range in moments plot

- use flat prior when request, or when have tNCS or twinning
- implementation is based on acentric case for Wilson prior

- correct Ice Rings statistic as we are working with merged data, unlike aimless

- introduce RNA/DNA reference curve

- bug fix to Optical Resolution routine

1.8.7, 1.8.8
- quick escape in case of failed anisotropy calculation

- update for gcc 4.7

- allow poorer data to be used

- use ML aniso values, corrected for Biso, to scale prior for truncate
- this should gain some of the speed back

- use ML aniso values, corrected for Biso, to scale data for twinning tests

- restrict resolution range for aniso to tidy plots

- simplify WilsonB class and plotting to reduce confusion

- include completeness test for data quality. Currently hardwired at I/sigI > 3 and above 85%
- more flexible Scattering class for calculation of the total scattering
- Completeness, tNCS, YorgoModis, ResoCorrel classes added for analysis

- make twinning test output more consistent

- use anisotopy in prior for truncate procedure
- no-aniso keyword now controls use of anisotropy in prior
- anisotropy corrected for twinning statistics, now not optional

- acentric tabulated data extended to h=10 to reduce discontinuity in plots
- Phil plot added to monitor output


  1. S. French and K. Wilson, Acta Cryst. A34, 517-525 (1978).
  2. T. O. Yeates, Acta Cryst. A44, 142-144 (1980).
  3. J. E. Padilla and T. O. Yeates, Acta Cryst. D59, 1124-1130 (2003).
  4. P. Zwartz, CCP4 Newsletter 42.
  5. Z. Dauter, Acta Cryst. D62, 867 (2006)
  6. P. Zwartz, Acta Cryst. D61, 1437 (2005)
  7. Schnieder, Acta Cryst. D58, 1772 (2002)
  8. B.C. Wang, Methods Enzymol. 115, 90 (1985)


Charles Ballard, Rutherford Appleton Laboratory.

Norman Stein, Daresbury Laboratory.