Dear Kevin,

It's good you tried genthreader to id the most suitable template. from here, try running a BLAST at NCBI database to locate pdb template sequence that have ideal percentage similarity. Hope this helps.

Sam.