Label distribution
Vitaly Selivanov
|
|
Label distribution
Vitaly Selivanov |
If substrates of cellular metabolism contain labeled atoms (as 13C isotopes), metabolic networks introduces them into the intermediates and products in accord with the internal metabolic fluxes. The distribution of labels contains information about the distribution of cellular metabolic fluxes, or metabolic flux profile. The label distribution could be assessed experimentally with NMR or GC/MS technique; analysis of these data requires tracing of huge amount of possible isotopic isomers (isotopomers) since a molecule of n carbon atoms could have 2n different isotopomeric forms. We developed a method of authomated construction of the respective huge system of differential equations, which simulates the dynamics of each possible isotopomer concentration in all metabolites of the following scheme.
This method is described in several publications:
The simulations, data fitting and statistical analysis of the obtained sets of parameters and fluxes could be performed ON LINE and there is a user guide, which explains how to work with this program. The C++ sources could be downloaded here.
This method was realized as a "Mathematica" (Wolfram Inc) program: fitting.nb.
This program first calculates total metabolite concentrations and fluxes for glycolytic and pentose phosphate pathways and then uses the linked program check1.exe to calculate concentrations of all possible isotopomers. This program was written in C, which makes these time-consuming calculations faster.
The present approach allows implementing in the isotopomer analysis the available "classical" enzyme kinetic information, which restricts ambiguity in the possible estimated sets of fluxes. The software, which accounts for the detailed kinetic mechanism of transketolase (TK) and transaldolase (TA) reactions, shown in the scheme below. The "Mathematica" program is in the file "rib-ot.nb" located in the archive
Bioinform2005Math. This contains also the executable "rrib.exe", which is linked to the main program; supplementary file "TK-scheme.nb", which shows the details of implementation of the isotope-exchange fluxes produced in TK and TA reactions. This program also has an algorithm for parameter optimization, which minimizes χ2 by coordinate descent, estimates the goodness of fit and formal standard errors for parameters. Development of new algorithm was necessary because the standard "Mathematica" algorithms for optimization do not work properly with differential equations because of falling to the area of parameters where the system becomes stiff and solution could not be obtained.
The supplementary file "TK-substrate.nb", simulating substrate accumulation in
TK reaction is located in the archive TK-substrate.
A C++ console application "coli530.exe" computes the distribution of all
isotopomers (though only mass isotopomers are included in the output)
accounting for, in addition to that present in the "Mathematica" version, the competition between
the TK substrates, Krebs cycle and anaplerotic reactions. It is located in the
archive
Bioinform2005-C++, also containing the scheme of the reaction pathway,
"reaction scheme.jpg"; the scheme of TK-reaction accounting for the substrate
competition implemented in "coli530.exe", "TK-compet.jpg"; the scheme for the
TA-reaction, "TA.jpg", and other files described in details in the user guide
"readme console.doc", also located in the same archive.
A "Windows" application ("wst602.exe") is located in the archive "windows", which contains also all the files necessary for the correct execution of "wst602.exe" and the user guide with the name "readme windows.doc". This windows
application uses the same format of files with experimental data "exper.dat" and
parameters "nv.txt" as the console application.
The archive "bioinf-Marin", corresponding to the publication: Bioinformatics, 2006, 22, 2806-2812 represents more complete
tool for metabolic flux profile evaluation from mass isotopomer
distribution. It analyses the metabolic network consisted of glucose input,
glycolysis, glycogen metabolism, Pentose Phosphate Pathway with detailed scheme
for isotope-exchange fluxes created by transketolase and transaldolase catalysis
accounting for competition between the substrates, Krebs cycle, anaplerotic
reactions, and various fluxes connecting this part with whole metabolic
system. In the present example it is customized for rat liver cell metabolism
and simulates different experiments where labeled glucose or labeled lactate
were used. It also has a procedure for χ2 minimization based on Simulated Annealing algorithm for global optimization, which uses Powell's method in a local area. It also analyses the sensitivity to parameters and evaluates formal standard deviations for the adjusted parameters. The brief user guide "readme.doc" describes the most essential details mecessary for using this tool. A small enterprise, Noraybio, made a grafical interface for this program in Windows, which could be found here.
NEW: Mitochondrial Respiration Model of mitochondrial respiration
and ROS production is a tool for the analysis of experimental data
related with mitochondrial respiration and coupled reactive oxygen species (ROS)
production. It describes the redox state of the complex, which is composed of
concentrations of 400 different combinations of oxidized or reduced states of
electron transporters: cytochrome Bh, cytochrome Bl, cytochrome C1, Fe-S complex
of Rieske protein, and ubiquinone/semiquinone/ubiquinol bound on cytosolic or
matrix sides of the complex.
Click here to download the model.
Research Groups
of the Department
V.A. Selivanov, L.E. Meshalkina, O.N. Solovjeva, P.W. Kuchel, A.Ramos-Montoya, G.A. Kochetov, P.W.N. Lee, M. Cascante. Rapid simulation and analysis of isotopomer distributions using constraints based on enzyme mechanisms: an example from HT29 cancer cells. Bioinformatics, 2005, 21, 3558-3564.
V. A. Selivanov, S. Marin, P.W.N. Lee, M. Cascante. Software for dynamic analysis of tracer-based metabolomic data: estimation of metabolic fluxes and their statistical analysis.
Bioinformatics, 2006, 22, 2806-2812.
Department of Biochemistry
and Molecular Biology
Last update: July, 14th 2008