FSEARCH (CCP4: Supported Program)
NAME
fsearch - A program to perform up to 6 dimensional molecular
(envelope) replacement search.
SYNOPSIS
fsearch hklin foo_in.mtz [flmin foo_in.txt]
[xyzinfoo_in.pdb] [mapin foo_in.map]
[Key-worded input]
DESCRIPTION
FSEARCH can be used for a molecular-replacement (up to 6 dimensional) solution
given a predetermined envelope from any source, provided that the envelope
can be converted to the standard CCP4 map format or expressed in terms
of spherical harmonics or generated from a pdb file.
INPUT AND OUTPUT FILES
Input
-
HKLIN
-
(compulsory)
-
Input MTZ file. This should contain the conventional (CCP4) asymmetric
unit of data (must be *sorted* on H K L). See the LABIN keyword for columns
used.
-
-
(Envelope input mode; ONLY one from the following three)
-
FLMIN
-
Spherical Harmonics File. Free format: L M Flm(re) Flm(im).
-
XYZIN
-
Coordinates File (Brookhaven format).
-
MAPIN
-
Map File - one asymmetric unit only (standard CCP4 map file format).
Output
-
Results (up to 200 solutions sorted by R-factor) are shown in the log
file.
KEYWORDED INPUT
The various data control lines are identified by keywords. Only the first
4 characters of a keyword are significant. The cards can be in any order.
Numbers in [ ] are optional and can be left out. The only compulsory command
is LABIN. The available keywords are:
TITLE, LABIN,
GRID, RESOLUTION,
SIGCUT, RFILTER,
ORTH, CHKOVERLAP,
ALPHA, BETA,
GAMMA, XRANGE,
YRANGE, ZRANGE
TITLE <title string>
Up to 80 character title.
LABIN FP=... SIGFP=...
(Compulsory.)
E.g.
FP=F SIGFP=SIGF.
GRID <nx> <ny> <nz>
Number of sampling divisions along whole cell edge. For all space-groups,
NX,NY,NZ must be even and must have no prime factors greater than 19. Default
values: nx = a, ny = b and nz = c (nearest integer in Angstrom).
RESOLUTION <rmin>
High resolution cutoff in Angstrom (default value = 10).
SIGCUT <Nsig>
Reflections will be excluded if F < Nsig*SIGF. Default: no reflections
are excluded.
RFILTER <rf>
Solutions with R-factor < rf will be written to the log file. Default:
rf = 0.60.
ORTH <code>
Orthonormalization code.
= 1 orthogonal x y z along a, c*xa, c*
(Brookhaven, default)
= 2 b, a*xb, a*
= 3 c, b*xc, b*
= 4 a+b, c*x(a+b), c*
= 5 a*, cxa*, c (Rollett)
CHKOVERLAP
Check packing clashes (% of the whole unit cell). Note: this can be *very
time consuming*, therefore not recommended for the initial search. Default:
no checking.
ALPHA <min> <max> <step>
Search range and step on Eulerian angle alpha (in degrees). Default: 0
180 3.
BETA <min> <max> <step>
Search range and step on Eulerian angle beta (in degrees). Default: 0
180 3.
GAMMA <min> <max> <step>
Search range and step on Eulerian angle gamma (in degrees). Default: 0
180 3.
XRANGE <min> <max> <step>
Search range and step on X (in Angstroms). Default: 0 a/2 1.
YRANGE <min> <max> <step>
Search range and step on Y (in Angstroms). Default: 0 b/2 1.
ZRANGE <min> <max> <step>
Search range and step on Z (in Angstroms). Default: 0 c/2 1.
PRINTER OUTPUT
Results (up to 200 solutions sorted by R-factor) are shown in the log file.
EXAMPLES
Unix examples script found in $CEXAM/unix/runnable/
fsearch.exam
Non-runnable example script
fsearch hklin pfsodb.mtz xyzin sod3.pdb <<eof
TITL SOD
# GRID 48 52 150
LABI FP=FP SIGFP=SIGFP
ALPHA 0 0 1
BETA 0 0 1
GAMMA 0 30 3
XRANGE 0 18 3
YRANGE 0 25 1
ZRANGE 0 58 1
RESO 6
eof
PARALLELISED VERSION
Code for an MPI-parallelised version of FSEARCH is distributed as
fsearch_mpi.f This is not integrated into the CCP4 installation,
and needs to be built separately. You will need:
- An MPI compiler, such as
MPICH or LAM.
- The link line will need to include to the MPI library as well as the
usual CCP4 libraries (possible names libmpich.a, libmpif.a, libfmpi.a)
- The source code fsearch_mpi.f includes the system file "mpif.h". This may
need to be adjusted.
The program distributes the search over Eulerian angles alpha and beta
over the available processors (with one processor reserved as a master
node). By default, 3 processors are used for the beta search for each
trial alpha angle. The search over alpha is then split over how ever many
sets of 3 processors there are. If the total number of processors (not
including the master node) is not a multiple of 3, then the last 1 or 2
processors are not used. The number of processors for the beta search
can be changed from the default value of 3 by the keyword PBET (e.g. "PBET 5").
REFERENCES
- Hao, Q. (2001), Acta Cryst. D57 1410-1414. "Phasing from an Envelope".
- Q. Liu, A. J. Weaver, T. Xiang, D. J. Thiel and Q. Hao,
(2003) Acta Cryst. D59 1016-1019.
"Low-resolution molecular replacement using a six-dimensional search"
AUTHOR
Quan Hao, Cornell University.
Email: qh22@cornell.edu.