This is a fork of the SBMLR package originally written by Tom Radivoyevitch(Case western University) used to import SBML models in R. 

The original package was very strict in terms of arrangement of the XML and rarely worked properly with most models available at Bioc.

In this fork, the parser for the most part has been rewritten. Some leeway has been given for parsing dirty XML's too.

The package can also be used to simulate Biological models within the R environment and perform global sensitivity analysis by computing 
the partial rank correlations of the different parameters of the model.

--Dependencies-----

libxml2 - XML tree creaetor. This can be installed by typing 'library("XML")' without the single quotes within R.
odesolve - Livermore ODE solver. This can be installed by tying 'library("odesolve")' without the single quotes from within R.

To perform sensitivity analysis, you need an additional package 'library("sensitivity")'. This can be installed from within R the same way as the previous packages.
----Installation---

Note: These Instructions apply to linux, windows and Mac operating systems.

1) Launch R and install the above packages.

2) close R and then at the prompt, type ' R CMD INSTALL /path/to/SBMLR/download SBMLR'

so for example, if the package was in /home/user/Downloads/SBMLR, you would type

'R CMD INSTALL /home/user/Downloads/SBMLR SBMLR

3) If all goes well, It will give you a success loading message.

4) Launch R. Type 'library(SBMLR)' and It should successfully load the package. You are now free to use the different functions.


----------------

Quick Start:

1) Install the dependencies and the library. Call the library within R using 'library(SBMLR)'.

2) Note down the path to your model and your Sobol Set for the parameters. If you are not using a randomized algorithm like Sobol to generate model inputs, simple type
	output <- load_simulate(1, path_to_model, time_series)
	The output is just raw data. The ideal way to use it to then bind it to a data frame like this. 
	Usable_output <- data.frame(output)

3) If you have a table of model inputs which you want to run, This can be imported easily. All you need to do is type,
	output_set <- load_simulate(path_to_sets, path_to_model, time_series)
	The output_set obtained is a list. Access the results like you access any list of simulation results. eg. output_set[[index]]

4) Look at the additional docs for using some of the individual functions if you want to calculate 'Area under the curve', GSA etc. 

Note: I recommend you use the load_simulate() function initially until you are comfortable using the package. After that, you can tweak and perform operations separately by calling the operations you want at the time you wish. The load_simulate() is just a wrapper which sets many things for you automatically.

Functions available in SBMLR

readSBML('path/to/model'):

Argument - SBML model file(with extension .xml)
Returns - Equivalent R model object file.

example
model = readSBML('/home/user/model.xml')

saveSBML(R model object, 'path/to/save/model.xml'):

Arguments - Path/to/save - where you want to save the file
	    R model object - the R object you want to save.

Returns - TRUE if successful. False if not.

example

saveSBML(model, '/home/user/some_name.xml')

simulate(model, time series):

Arguments - model: The R model object created using the readSBML function.
	    time_series: The number of minutes you want to perform the simulation. This should be a vector.

Returns - Simulated R object file.

example

simulated = simulate(model, seq(1,60,1))			#here time is a vector sequence from 1 to 60

-------------------------

Thats all for now.
---------------------
Readme written by vishak, Msc Computer Science student, Edinburgh University.