The files in this directory walks you through an example to help you
run the antibody modeler mode. Before proceeding you need to make sure
that you have done the following:
1. Updated your code to make sure that you have the latest
	 developments
2. Make sure that you check out the "antibody" directory in the trunk
3. Read the "antibody_readme.txt" file in the "antibody" directory and
	 make sure that you follow all the instructions

After all the scripts and supporting files are in place untar the
"antibody_examples.tar" file. A description of the files in tar
archieve is provided:
a. grafting_input <directory>		This contains all the fasta sequence
																files for the query. 
   i.   launcher.bash						A bash file to invoke the master
																wrapper script. The script takes the
																fasta files as an input and outputs a
																framework with everything except the
																CDR-H3. It also generates the fragment
																files for H3 modeling. It also creates
																a build directory with a condor script
																to actually build <nstruct> number of
																homology models
	 ii.  query_h.fasta           Fasta file containing the heavy chain
												   			sequence
	 iii. query_l.fasta           Fasta file containing the light chain sequence 
   iv.  rampaths.txt            Paths for various files needed by scripts 
	 v.   utilities.txt           Supporting file
b. build_input <directory>      This contains the files needed for
																modeling the CDR H3. These files are
																the output files generated by
																"launcher.bash". 
	 i.   1xyz_build_loops.bash   Bash script that invokes rosetta to
															  actually model the CDR-H3 loops. This
																is the most time consuming part of the
																protocol. For simplicity the nstruct
																is set to 1. However, for a normal run
																it should be 2000. The file created by
																"laucher.bash" actually has nstruct
																2000. I manually changed it to 1 for
																the purposes of this example.
   ii.  aaFR02_09_05.200_v1_3   Standard rosetta 9-mer fragment file
   iii. aaFR02_03_05.200_v1_3   Rosetta 3-mer fragment file appended
																with antibody fragments
   iv.  H3_CTERM								Special fragment file containing H3
																base fragments 
   v.   hfr.pdb									Template structure from which heavy
																chain of query has been obtained 
	 vi.  lfr.pdb                 Template structure from which light
																chain of query has been obtained   
   vii. paths.txt               Standard rosetta paths file. Make sure
																this matches with your paths. Once
																again this should be set if you have
																followed instructions in "antibody_readme.txt"
   viii.FR02.pdb                The output file of "launcher.bash"
																which has the all non-H3 CDRs grafted
																onto the framework regions. It
																contains the CDR-H3 region with the
																coordinates blanked out.
c. build_output <directory>     Files generated by the "1xyz_build_loops.bash" 
	 i.  	aaFR02_0001.pdb					Output decoy with ab initio built
																CDR-H3. The non-H3 CDRs alongwith the
																relative orientation of the light and
																heavy chains has been optimized
   ii.  FR02_dummy_loop_01_0001 A fragment generated for the CDR-H3
																loop modeled plus one flanking residue
																on either side. The three columns are
																for the phi-psi-omega angles for each
																residue. This is helpful in recreating
																the loop using other command line
																options like "refine_loop"
   iii. aaFR02.fasc             Scorefile with special columns:
				a. rmsg                 Global rmsd of CDR H3 if native was
																specified, otherwise -1.00
				b. hydrphbc             A hydrophobic score adapted from
																Friesner's long loop modeling paper
				c. Brms                 Global rmsd of the CDR H3 base region
				d. Rrms                 Global rmsd of the CDR H3 non-base
															  region
				e. Gap                  Separation at the loop
																cutpoint. Useful in detecting broken
																loops. Has to be less	than 1.9 Ang
        f. H3                   Binary variable is true if decoy
																satisfies Shirai rules
			  e. N--O                 Distances between the backbone oxygen of
																the (n-2)th residue and the NEl atom
																of the (n+l)th Trp side chain. (Shirai)
				f. HH										Number of hydrogen bonds in the
																CDR-H3. This is purely based on a
																distance based criteria and does not
																use Rosetta's sophisticated hydrogen
																bond stuff. Helpful in detecting H
																ladder formation
			  e. St                   Checks if Stem follows Shirai's rule
																for extra bulge.

Make sure that launcher.bash and rampaths.txt are modified to match
your paths. You should have completed this already if you have read
"antibody_readme.txt" in the "antibody" directory. You can merely copy
the rampaths.txt from "antibody/scripts/rampaths.txt" to overwrite
this local copy.

Once you have modified the "launcher.bash", after the run finishes,
you should have a build directory. There are thousands of files in
there, but then I have placed the key files in the "build_input"
directory. You only need these files to be able to build the CDR-H3
into the antibody scaffold. 

Finally, on execution of the "1xyz_build_loops.bash" script, we
actually generate the final homology models. We generally build 2000
of this (in this example, we build just one) and choose the top ten. 
