MCSS Docking

Hi all,

                 I want to use MCSS to predict new inhibitor molecules using fragments availbale. I have question:

1.  I want to use those framents obtained from already known inhibitors , is it reasonable?

2. when i try docking some fragments, all fragments were found to be docked outside the defined sphere, please tell me where i am going wrong in protocol?

3. After generating few possible fragments, how can i connect them to generate complete molecules (whats methods can i use)..

You advice will be helpful to my work...

Thank you

with regards,

vamsi