+---------------------------------------------------------------------+

| ____ __ __ ___________ ______ |

| / __ \/ /_ __ __/ /____ / ____/ ___// ____/ |

| / /_/ / __ \/ / / / // __ \/ / \__ \/ /_ |

| / ____/ / / / /_/ / // /_/ / /___ ___/ / __/ |

| /_/ /_/ /_/\__, /_/ \____/\____//____/_/ |

| /____/ |

| |

| http://compbio.mit.edu/PhyloCSF |

+---------------------------------------------------------------------+

SOURCE DISTRIBUTION

Please see our web site for more information about PhyloCSF and how to use it.

PhyloCSF is written in OCaml. This source distribution is divided into a

library, CamlPaml, containing generic infrastructure for phylogenetic rate

models, and a program implementing the PhyloCSF-specific models and driver

program. With additional development, the CamlPaml library will eventually be

released as a separate entity, but for now it is just part of this

distribution.

Here are the steps to build the source:

[1] Install the OCaml package manager OPAM

http://opam.ocamlpro.com/doc/Quick_Install.html

This will install a minimal OCaml toolchain (if not already present on your

system) and the OPAM package manager. Once it's installed, initialize your

shell environment by running: eval $(opam config env)

[2] Install dependencies using OPAM

opam install batteries ocaml+twt gsl

The gsl package may require you to separately install the GNU Scientific

Library (GSL) from: http://www.gnu.org/software/gsl/

On Debian/Ubuntu systems GSL is available in the 'libgsl0-dev' package, and on

Mac OS X, 'gsl' is available in both Homebrew and MacPorts.

[3] Compile PhyloCSF

Now just run 'make' in this directory. This will build and install the

CamlPaml library, then compile the PhyloCSF executable and copy it to

this directory. You can then use the ./PhyloCSF dispatch script like the

examples found on our website, e.g.

./PhyloCSF 12flies PhyloCSF_Examples/tal-AA.fa

PhyloCSF should be usable in this way by anyone on the machine/cluster with rx

permissions; they don't have to set up OPAM or anything else. PhyloCSF can

be started from any working directory.

If you want to 'install' PhyloCSF somewhere, you must transplant PhyloCSF,

PhyloCSF.ARCH (where e.g. ARCH=Linux.x86_64), and the PhyloCSF_Parameters

directory to the desired location.