== Prep work ==
As listed in the dependancies section below, you need
 * Python 2.7.

== Included external applications ==
The following applications are included and will compile and be used automatically, 
however you should try installing them yourself because you can get substantial
performance increases by choosing makefile which are more appropriate for your system.
* FastTree (>=2.1.4) http://www.microbesonline.org/fasttree/,
* FASTML (2.02) http://fastml.tau.ac.il/,
* RAxML (>=7.2.8, the full version) http://sco.h-its.org/exelixis/software.html, and

Finally, the raxmlHPC binary should be named "raxmlHPC".  If you have a modern CPU you 
can get vastly increased performance by installing RAxML yourself and selecting
the most appropriate makefile.

== Installing from a git checkout ==
$ autoreconf -i
$ ./configure
$ su -
$ make
$ make install
$ cd python
$ python setup.py install

== Installing from a released source tarball ==

$ tar xzvf gubbins-0.1.tar.gz
$ cd gubbins-0.1
$ ./configure
$ make
$ su -
$ make install
$ cd python
$ python setup.py install


To run the script with all of the default bundled executables.
$ cd python
$ scripts/run_gubbins my_alignment.fa


The main executable should be in ./src/gubbins and in the system-wide directory
for locally compiled applications.  On Linux this is /usr/local/bin.

On Linux, you should ensure that /usr/local/bin is in your $PATH and that
/usr/local/lib (or /usr/local/lib64 on 64bit machines) are in $LD_LIBRARY_PATH.

== Installing into a specific user directory ==

It is common to want to build and install an application so that it is not
system-wide and is only available to your user.  On a Linux system this is
generally achieved by installing into the .local directory in your home
directory.

$ ./configure --prefix=~/.local
$ make
$ make install

On Linux, you should ensure that ~/.local/bin is in your $PATH and that
~/.local/lib (or ~/.local/lib64 on 64bit machines) are in $LD_LIBRARY_PATH.

If you are in the situation where your home directory is mounted over NFS (or
similar) and shared between multiple Linux distros (for example, 32bit Debian on
a cluster and 64bit Ubuntu on a desktop) then installing per-user in this manner
is probably not a good idea.

== Dependancies ==

If these applications are not available in your PATH, less efficient ones will be used:
FastTree (>=2.1.4) http://www.microbesonline.org/fasttree/
RAxML (>=7.2.8, the full version) http://sco.h-its.org/exelixis/software.html
FASTML (2.02) http://fastml.tau.ac.il/

== Python ==
Python 2.7 is required

These Python packages must be in your PYTHONPATH or will be automatically 
installed:
Biopython (>=1.59)
DendroPy  (>=3.11.1)
ReportLab (>= 2.5)

== Running the Python wrapper ==

There is a Python wrapper script that automates some of the preparatory work
that is necessary to run Gubbins.

To run the application:
run_gubbbins.py my_alignment.aln