Readme.FIRST

** Build procedure for BBNet and GBNet **
Extract all source files into one folder. To build simply type:

make

in command line. It will automatically start building the executable files.
After that, if you want to clean your folder, simply type:

make clean

in command line. It will delete all object files but keep the executable.
In case you modify the source files, you may also want to clean the objects first and then "make"
to make sure your updates are reflected in executables.

** Verbose Mode **
The default setting will produce a verbose version of BBNet which outputs each step of the 
Bayesian network structure change. To switch off this behavior, you can do it like this:

vi globals.h

to edit globals.h. In the file, find "#define VERBOSE" and then turn it off by applying "//" to the 
start of the line, like this: "// #define VERBOSE". This tells the preprocessor to bypass this line 
so it won't be compiled. In command line, type "make" to re-build BBNet. 
Therefore, all outputs to screen will be muted.

** Functional Depth **
The default functional depth varies from 0.05 to 0.95 as described in Beer's seed paper. 
We extened it to add additonal depth from 0.01 to 0.04 because we found that in higher organisms, 
binding motifs tend to be more degenerate than in lower ones. To switch on this additional depth, 
you need to do some manual work as follows: 

vi globals.h

to edit globals.h. In the file, find "#define YY1FUNC" and then delete the preceding "//". 
In command line, type "make" to re-build BBNet and GBNet. 

** Procedure to run **
A typical procedure to get data prepared and the program running is like this:

func -> bayescor -> bbnet or gbnet

func will prepare the functional depth files; bayescor calculates the Bayesian score for each 
individual motif; bbnet or gbnet run the main programs of BBNet or GBNet.