CCL: MonteCarlo search

Dear  CCLers,
 I am working with a set of flexible molecules containing  a ciclopentyloxy
 residue and I am  getting thousands of conformers from MonteCarlo search.
 By cluster analysis these conformers  are grouped in two or three big families
  instead of being uniformely grouped. Is this something that could happen with
 very flexible molecules?
 Shall I take into consideration only the lead of the most populated groups or
 all conformers selected by cluster analysis?
 Thank you very much for your answers,
 Dr. Eleonora Lo Presti
 Dip. Scienze Farmaceutiche - Facolta' di Farmacia
 Universita' di Genova - Viale Benedetto XV, 3 - 16145 Genova Italy
 lopresti &$at$&