Use NMR structures for comparative modelling

[Chyh-Chong Chuang]

        I am curious how to use the NMR structures as the base stucture for
comparative modelling using the avalible commecial packages, Homology,
Composer,.. As I know, the NMR structures are low resolution. And usually
the side chains are not well defined. Is it reasonable to use the average
structure as the base structure? Is it reasonable to trnsfer the side
chains' orientation to the conserved residues.
        Could anyone give me some suggest/comment about this practical
question? Thanks all.

ccchuang