Format of files:

models:     all-atom in PDB format.


template list:  1st column: template; 
                2nd SP^3 Z-score; 
                3rd sequence ID to target.
                Templates like 3pblac,3rzeac,3oduac are structures with the engineered domain cut away.  

SP^3 alignments:  In MODELLER format

ligand screening data: 1st column rank #
                       2nd compound ID
                       3rd compound virtual screening raw score of continuous Tanimoto coefficient with 
                           normalized ligand profile 
                       4th compound virtual screening Z-score 
ligand profile data:  This is a FINDSITE  pocket file, last line contains the normalized ligand 1024 bit 
                      fingerprint profile. Users can use it in continuous Tanimoto coefficient calculation 
                      against their own compound library. A program of doing screening using FINDSITE pocket
                      can be find at 

                      http://cssb2.biology.gatech.edu/gpcr2011/findsite_screenG

                      Refer to FINDSITE manual for more details:

                      http://cssb2.biology.gatech.edu/genome/findsite_manual-1.0.pdf