Integrated software (collaboration: Minor Lab, U.Va.) /
Links: HKL
Crystallographic data integration
Data scaling and merging assessment
Phasing /
Links: MLPHARE at CCP4
Model building
- Applications of Maximum Entropy principle
to modeling torsion angle probability distribution in proteins
- Stereochemical Rules for Connecting Disjoint Protein Fragments
- Bayesian statistical studies of the Ramachandran distribution
Posters:
- New approach to protein model building
- Stereochemical rules for connecting protein fragments
Symmetric Real Spherical Harmonics
- An efficient routine for computing symmetric real spherical harmonics for high orders
of expansion
Crystallographic Fast Fourier Transform
We have developed a new approach to crystallographic Fast Fourier
Transform (FFT).
It resulted in algorithms for all 230 space groups.
The presented algorithm allows
to reduce both
computation time and memory usage
by a factor approximately equal
to the number
of symmetry operators
in the crystallographic group.
The solution
is ultimate: it has
reached the theoretical limit of computational complexity and it
is highly efficient for current computer architecture.
We are currently
preparing the first release of the KRFFT software library.
(click here for more details)
The algorithms have been described
in the following papers:
- p3
symmetry
- One-step
symmetry reduction
- Centered
lattices
- FFT-asymmetric units in the reciprocal space
- Recursive symmetry reduction
|
