Dear all I expect this is more a question for the developers, but perhaps other users have relevant experiences to share. With 9v5 we have the very useful Z scores from a normalized DOPE assessment. The manual says that +ve scores are bad, and that anything less than -1 or so is 'native-like'. Presumably the best achievable Z score will depend very much on the available templates - we should be able to do much better with 70% identical templates than with 30% identical templates. Has anyone got a feel for what kind of figures we should be happy with (and be able to quote in a paper as supporting the quality of our model) in different ranges of template % sequence identity? Thanks in advance Dan -- Dr Daniel John Rigden Tel:(+44) 151 795 4467 School of Biological Sciences FAX:(+44) 151 795 4406 Room 101, Biosciences Building University of Liverpool Crown St., Liverpool L69 7ZB, U.K.