CCL: Estimating applicability of fingerprint model

["Iain Wallace" <iain.m.wallace{=}gmail.com>]

 Hi all,

I have built a model to classify compounds into two classes using
 the Pipeline pilot Bayesian fingerprint classifier (ECFP_4 Fingerprints).  
 I was wondering if anyone has any experience on how to estimate how well the
 model I have built will transfer to other libraries? I know that I should only
 apply the model to compounds drawn from a similar distribution, but I have no
 idea how to what steps I should take to ensure that this criteria is met.
 
For instance, I would like to score the Zinc database (all commercially
 available compounds) to find new interesting molecules, and I am wondering if
 anyone has any tips on problems I should look out for,

Thanks for alot
 for any advice 
 
Iain