Program:
Garli v2.01

Comments:
Velvet manipulates de Bruijn graphs for genomic sequence assembly. The program represents a new approach to assembly that can leverage very short reads in combination with read pairs to produce useful assemblies.

Source:
https://www.nescent.org/wg_garli/Main_Page

Cite:
Zwickl, D. J., 2006. Genetic algorithm approaches for the phylogenetic analysis of large biological sequence datasets under the maximum likelihood criterion. Ph.D. dissertation, The University of Texas at Austin.

Compilation notes (serial):
1. First compile NCL library as per Garli INSTALL instructions
	From ncl root directory (within Garli):
	$ env CPPFLAGS=-DNCL_CONST_FUNCS ./configure
	$ make
	$ make install
2. Then compile Garli (serial execution)
	From Garli root directory:
	$ ./configure --prefix=/home/sgi/GARLI201serial/bin --with-ncl
	$ make
	$ make install

Compilation notes (parallel):
1. First compile NCL library as per Garli INSTALL instructions
	From ncl root directory (within Garli):
	$ env CPPFLAGS=-DNCL_CONST_FUNCS ./configure
	$ make
	$ make install
2. Then compile Garli
	From Garli root directory:
	$ env CC=/opt/sgi/mpt/mpt-2.07/bin/mpicc CXX=/opt/sgi/mpt/mpt-2.07/bin/mpicxx ./configure --prefix=/home/sgi/GARLI20parallel/bin --with-ncl --enable-mpi
	$ make
	$ make install

The argumentet --enable-mpi:
		(build the MPI run distributing version.  It should automatically determine
		the correct MPI compiler script name (e.g. mpiCC) but if the underlying compiler
		is not gcc (try "mpiCC --version" to check) you should also set CC and CXX to the
		corresponding compiler type (e.g., CC=icc and CXX=icpc) to get the right compilation flags)--enable-mpi

Sudo-parallel Garli:
We wrote a Python script, sudoParallelGarli.py (available at http://www.ib.usp.br/grant/anfibios/homePage.html), to emulate Garli parallelization using a Garli executable compiled in serial mode.