Max Planck Institute for Molecular Genetics

Kinetic Modelling Tools

TIde

Tide is a tool for the automatic identification of optimal drug targets in kinetic models based on ordinary differential equations. Give a model in the popular SBML format it will identify promising drug targets for different effective modifier concentrations.

Downloads

TIde 1.2.1
SBOS is a Linux live DVD containing a working installation of TIde. Just download the ISO image, burn it, and boot your computer from this DVD.
Windows installer for all required libraries. Note that TIde has to be downloaded separately.

Requirements

Python <3.0, ≥2.5, SciPy ≥0.5, NumPy (Numeric and NumArray might be needed depending on the SciPy version), semanticSBML, libSBML 3.x.

Optional

PyX (for pdf output)
Copasi 4.4, SOSlib 1.6 (as different ODE solvers)
ParallelPython 1.5.3 (for parallel computation, not documented).

Install

First, you have to install all the required tools and libraries from the Requirements section. Second, extract the tarball into a directory of your choice. Then download the Systems Biology Ontology into the TIde directory. You will now be able to do some configurations in the file global_options. If you are running Linux, then you will be able to set the solvertype to c and if you have additionally Copasi or SOSlib installed, then the you will also be able to set it to copasi or soslib. If you do not want to use any external simulation software, set the solver to scipy. In case you have ParallelPython installed, you can set the use_parallel option to 1 and add available servers to the servers file. The lines in the file are statements of the form ip,computer_name,number_of_cpus_to_be_used.

How to use

For the analysis of a single file you have to call python model.py -f path_to_model in the TIde main directory including the options

option	meaning	example	optional/mandatory	default
-a	Observable, e.g. species_1 which is accumulating in the pathological state	species_1	mandatory
-s	Model is a signalling pathway		optional
-i	Different effective inhibitor concentrations	0.1,1,10	optional	0.1
-g	Use only certain reactions as possible modification targets	reaction_0,reaction_1	optional	all
-l	Use up to this many modifications simultaneously	2	optional	1
-t	Do the computation by a simulated titration		optional
-e	Estimate effective inhibitor concentrations to achieve a reduction of the aim value to this factor.	0.1	optional

After all computations have been performed you can call different output methods, e.g. html.py, htmllist.py, or pdfplot.py. The first argument to the tools is the name of the output file which has to be placed in the folder named after the model which was created in the previous step. E.g. python html.py models/simple_signalling_model/sdh will create the file models/simple_signalling_model/sdh.html. The options for html.py are

option	meaning	example	optional/mandatory	default
-i	Ignore lines containing the strings or lacking the strings, e.g. show only combinations including reaction_3 but ignore entries with the effective inhibitor concentration of 0.375.	^reaction_3,0.375	optional
-s	Ignore control/response coefficients in the output.		optional

while the options for pdfplot.py are

option	meaning	example	optional/mandatory	default
-r	Range of the x axis	1,100	optional
-x	Take only curves whichs name matches this regular expression	".(1\|5).n$"	optional	.*

Another interesting script for analysing data of a 2-dimensional scan for good inhibition targets is sdhanalysis.py. With its help synergisms and antagonisms in the inhibition results can be revealed. The command line options for this script are

option	meaning	example	optional/mandatory	default
-t	Threshold how much stronger a synergism has to be compared to the sum of single effects	5	optional	1
-v	Verbosity	1	optional	0

Questions?

You can contact the author of this software under marvin!schulz#biologie!hu-berlin!de if you are smart enough to figure out the real email address.