Are you talking about any amino acid sequences or some particular ones
that are in the database? As far as I know, no groups can claim they
can produce the protein structure (as compared to X-ray structure say)
from the primary sequence. I know some groups are working on it
that is to predict the 3-D structure "a priori". And, they are thinking
of patenting it. I know of a CHARMM package but I am not sure whether or not
it is free.