Enrico,
At the time it was written, the RMS code was tested against other tools
and confirmed to produce equivalent numbers. Our superposition
algorithm is implemented using center of geometry translation followed
by an iterative three-way Jacobi rotation of the correlation tensor to
shrink the o
Joel,
> align 1kjg & i. 1-99 & n. ca, 1kjh & i. 1-99 & n. ca
pair_fit 1kjg///1-99/ca, 1kjh///1-99/ca
pair_fit will use all atoms, whereas align will only use atoms that are
matched with respect to a dynamic-programming sequence alignment.
Cheers,
Warren
--
Warren L. DeLano, Ph.D.
you can set the number of cycles of refinement on the command line.
change this:
> align 1kjg & i. 1-99 & n. ca, 1kjh & i. 1-99 & n. ca
to this:
> align 1kjg & i. 1-99 & n. ca, 1kjh & i. 1-99 & n. ca, 0, 0
I forget what the first 0 means; the second means "don't refine." It'll
use all matching
