What can CCP4@DL contribute to BioXHIT?
First a list of possibilities:
BIOXHIT was considered to be split in to 3 main sections
1. Crystallisation.
Low-cost LIMS ("Mole") being developed. Biologists in SBL are gaining
experience with Lisa, and will help to trial Mole.
Mole could be distributed to external groups (under CCP4 banner?)
and/or made available to synchrotron users?
Although both commercial (Nautilus) and academic (Xtrack) LIMS are
being developed within SPINE, there is still a niche for a
lightweight LIMS compatible with the CCP4 suite.
2. Data collection
Core development of MOSFLM. Support for Harry if CCP4 grant
funding fails (BMS committee meets January 23/24 2003).
This possibility has been raised, but is presumably intended
to be under Andrew who is separate partner.
Extension of DNA to partner sites. DNA is currently being extended
as part of e-HTPX. It might be possible to role this out to
other BioXHIT partners, though current DNA developers are not keen
to muddy the waters with extra partners.
3. Structure solution.
This is clearly our main area, and our strength is in providing
tools and infrastructure to enable application of algorithms.
With the current emphasis on data management, an important
advance here is the development of the ccp4i "database". Pete is
beginning to develop this
(See: http://www.ccp4.ac.uk/peter/dbhandler_specification.html
and will provide a more detailed case. The overall aims would
be to expand the ccp4i database within the CCP4 suite, and also
to enable communications with external applications, external databases,
LIMS and automated procedures. From the point of view of BioXHIT,
it should be possible to communicate between ccp4i Db and local
site databases.
Longer term aims would be to enable tracking in ccp4i, and
include basic integrity checking / data validation. These assume
ccp4i db goes beyond the current simple project history database.
We will also shortly have a greatly expanded computational toolkit,
based on the new libraries (DL libraries, MMDB and Clipper) and higher
level tools (Stuart's coord library, extensions of DL libraries). These
will be available to python via SWIG and/or Boost. This puts us in
a good position to develop automation scripts. These in turn will
involve development of python run scripts and a central "controller"
(these are the other components of the CCP4 architecture diagram).
And now a more selective response:
The most promising area is development of the ccp4i database and its
integration into other projects. There is certainly enough work here
for a PDA. Manpower could also be used for development of Mole, and
possibly for building on the new libraries.