Dear Donghai,
Unfortunately, CopySeq is currently not designed to run on exome
sequence data. The main reason for this is that the model parameter
estimation for the G+C bias and the read-depth variance estimation is
dependent on loci which are scattered all around the genome. At the
moment I am not planing to further develop CopySeq for exome
sequencing data since I am not working with this type of dataset.
Please have a look at one of these tools which might work on your
datasets:
http://seqanswers.com/wiki/Special:BrowseData/Bioinformatics_application?Biological_domain=Copy_number_estimation
Best wishes,
Sebastian