Inspired by  modeller's iterative tutorial I wounder to know more about MOULDING concept of modelling and its automatics implementation within the modelling scripts.  For instance I'm modelling of Rhodopsin-like membrane receptors from GPCR super-family with very slow sequence overlap with the closest templates (up to 20%) so I'm not sure both in automated-generated alignments as well as in the final models. How I could improve quality of both things using iterative concept?  Does anybody have experience with the GPCRs modeling ?