Most research papers I've read through regarding protein structures reference SCOP rather than CATH. Perhaps my sample size is too small. I'm confused because SCOP seems to be no longer in development. There is a site up for SCOP2 with latest news from 2014. On the other hand, CATH is being continually updated and provides recent versions of PDB files for consumption. Given this, my inclination would be to use CATH as the place to get domains and test against rather than a snapshot of SCOP data from some time ago. Any comments on this subject appreciated.