REMO mannual  
*******************************
REconstruct atomic MOdel(REMO) package
version 1.0


1. Introduction.

   REMO is a multifunctional program for constructing full-atom protein models from 
   C-alpha traces by optimizing the backbone hydrogen-bonding networks. It gives
   options to choose models which are close to a given template or have more hydrogen-
   bonds. It has also options to build hydrogen atoms and side chain heavy atoms. The 
   bond length and bond angle parameters are taken from CHARMM22.

   The package includes following files:

	1) REMO.pl : A perl script to setup and run REMO;
	2) FAMR    : The core excutable program to execute all REMO functions;
 	3) FAMRcomm: REMO parameters and potentials;
	4) BBdat   : The backbone isomer library;
	5) mkseq.pl: A perl script to for secondary structure distribution using PSIPRED.
	6) SRC     : A directory including source codes and executable files of FAMR and 
		     examples of input and output files. The source codes are in Fortran90
		     and the executable files include both 32bit and 64bit versions.

2. Installation.

   You need to first download the "remo.tar.bz2" file and then unpack the files by
   running "tar -xjf remo.tar.bz2".

   REMO consists of a perl program REMO.pl, a Fortran90 compiled program FAMR (include 
   32/64bit versions), and two parameter files of FAMRcomm and BBdat. Have all the four 
   files in the same folder, and change the $path0 in the REMO.pl to this folder. REMO 
   is then ready to use.

3. USAGE and options.

   There are following options in REMO:

	1) ./REMO.p1 0 xx.pdb seq.dat [template] <PRHB>
    	   Construct backbone structure from "xx.pdb" (only C-alpha atoms are used),
	   maximizing the hydrogen-bonding networks based on the secondary structure 
           distribution as given in "seq.dat" and PRHB (if provided), also draging the 
           xx.pdb closer to a template.
	output: xx.pdb.fix (model closest to "xx.pdb")
	        xx.pdb.h   (model closest to "template")

	2) ./REMO.p1 1 xx.pdb seq.dat seq.txt 
	    Build full length model according to sequence file "seq.txt"(Fasta format).
	output: xx.pdb.h (model closest to "xx.pdb")	
	
   NOTES:
	
	1) The example of input files can be found in @REMO/SRC/example folder. 
	2) Full-length structure from the template is require when including template. 
           And the "template" structure can be different to or same as "xx.pdb".
	3) REMO.pl has two modes:
	   normal mode: (set $mode="normal" in "REMO.pl") no other operations are needed.
  	   fast mode  : (set $mode="fast" in "REMO.pl") you need to first delete the "Hdd" 
			and "PRHB" files before you run a new protein at the same 
			directory, this provides double speed if you treat different 
			models of the same sequence.
	4) The residue number will be reordered from 1 in the output structure.

4. Advanced usage:

   REMO is a standalone program. But following options will improve the quality of 
   final models when running REMO.

	1) Accurate secondary structure distribution (file named as seq.dat in 
	   above options). You can generated secondary structure predictions
	   by "mkseq.pl"; but you need to install PSIPRED first and configure 
	   "mkseq.pl" accordingly. PSIPRED programs can be download from 
	   http://bioinf.cs.ucl.ac.uk/psipred

	2) Accurate Hydrogen Bonding network (PRHB, backbone only). PRHB contains 
	   five columns: series number (not used), "N", H-bond donor residue number, 
	   "O", H-bond acceptor residue number. The REMO itself can make this file 
	   based on the consensus between the input structure and the secondary
	   structure predictions (if existing) by default. But an additional H-Bond 
	   list file will help improve HB-scores.

	3) The REMO models initially contains side-chain atoms without optimization. 
	   But for further side-chain optimizations, you are recommended to install
	   Scwrl program from: http://dunbrack.fccc.edu/SCWRL3.php.

5. Contact.

   For question and comments, please contact: 
   Yunqi Li (yunqi@ciac.ac.cn) or 
   Yang Zhang (zhng@umich.edu).

6. Reference.

   Please cite following paper when you use the REMO program:

   Yunqi Li and Yang Zhang. REMO: a new protocol to generate full atomic protein 
   models from C-alpha traces by optimizing backbone hydrogen-bonding network. Proteins,
   76(3):665-676 (2009).