[ccp4bb]: fffear 1.9-beta (maximum likelihood version)

[Kevin Cowtan <cowtan@ysbl.york.ac.uk>]

***  For details on how to be removed from this list visit the  ***
***    CCP4 home page http://www.dl.ac.uk/CCP/CCP4/main.html    ***

PROGRAM ANNOUNCEMENT: fffear 1.9-beta

A beta-test release of the new maximum likelihood version of fffear is
available for ftp. This version uses similar assumptions to Tom
Terwilliger's maximum likelihood density modification, combined with the
quadratic search target of the original program, to implement a maximum
likelihood target for fragment searching.

The maximum likelihood target is much more effective when locating
fragments in poorly phased low resolution maps, and has been used to
locate helices in MIR maps truncated to 8.0A and SIR maps truncated to
6.0A resolution.

The construction of maximum likelihood targets currently involves
massive homology searches across large chunks of the pdb. As a result,
ML targets are only available for the empirical 9-residue helix fragment
at residues from 4-8 Angstroms resolution. It is hoped that further
analysis will allow the construction of synthetic ML targets for
arbitrary models, including MR models.


One problem with earlier versions of fffear was that in a few cases, the
search fragment would be fitted exclusively to the solvent rather than
the protein. This problem has not been explained, since I cannot
reproduce it using known structures and helical fragments. It is
possible that the ML target will resolve this issue, and I would be
pleased to receive any reports of success or failures in this area.
Alternatively, if any structures which demonstrated this behaviour with
a standard helix search fragment have since been solved, I would be very
happt to receive test data.


To pick up fffear 1.9-beta, pick up the code from
  ftp://ftp.ysbl.york.ac.uk/pub/ccp4/ffffear.tar.gz
It can be built using
  cd fffear/fffear_
  ./makefffear
  cd ../ffjoin_
  ./makeffjoin
The latest ccp4 libraries must be installed and the usual ccp4
environment initialised.

Kevin