constraining an serine triade

Hello from Marburg,

I want to do homology modelling on serine hydrolases with highly conserved fold and reasonable AA-identity. So the difficulty is to model the active site correctly with respect to the catalytic triade and the backbone nitrogens in the oxyanion hole.

These few amino acids should be overlayed as exact as possible.

Is there a straightforward way to contstrain modeller only for these higly conserved topology of the catalytic residues within one Angstroem, thus remaining the catalytic hydrogen bond network intact.

Any hints on possible protocolls appreciated.

Marco