Team:UFMG Brazil/modeling

From 2013.igem.org

(Difference between revisions)
(TMAO - Modeling)
(Model simulations)
 
(105 intermediate revisions not shown)
Line 1: Line 1:
{{Team:UFMG Brazil/barra}} {{carlos_teste}}
{{Team:UFMG Brazil/barra}} {{carlos_teste}}
 +
=Modeling TMAO system=
-
== TMAO - Modeling ==
+
==Labels and conventions==
 +
<html><b>C<sub>E</sub></b></html>: The enzymatic complex formed by TorT+TorS+TMAO, which reacts with TorR
-
'''Labels and conventions'''
+
<html><b>C</b></html>: intermediate component of the final enzymatic process
-
<math>C_{E}</math>
+
<html><b>P</b></html>: Product of the enzymatic reaction
-
: The enzymatic complex formed by TorT+TorS+TMAO, which reacts with TorR
+
<html><b>C<sub>TS</sub></b></html>: Complex formed by TorT and TorS
-
C: intermediate component of the final enzymatic process
+
<html><b>C<sub>TSM</sub></b></html>: Complex formed by TorT+TorS and one molecule of TMAO
-
P: product of the enzymatic reaction
+
<html><b>C<sub>TSM</sub></b></html>: Complex formed by TorT and TMAO
-
C_TS: Complex formed by TorT and TorS
+
==First Modeling==
-
 
+
-
C_TSM: Complex formed by TorT+TorS and one molecule of TMAO
+
-
C_TM: Complex formed by TorT and TMAO
+
-
 
+
-
'''First Modeling'''
+
This is the simplest modeling process to be considered in our study. The product of the enzymatic process is collapsed in only one reaction that was placed at equilibrium.
This is the simplest modeling process to be considered in our study. The product of the enzymatic process is collapsed in only one reaction that was placed at equilibrium.
-
[[File:eq1_tmao.png]]
+
[[File:eq1_tmao.png|300px|thumb|center]]
Imposing the equilibrium in the first reaction we have:
Imposing the equilibrium in the first reaction we have:
-
[[File:eq2_tmao.png]]
+
[[File:eq2_tmao.png|300px|thumb|center]]
-
The value of Ce is the initial enzyme concentration for the subsequent enzymatic reactions.
+
The value of <html>C<sub>E</sub></html> is the initial enzyme concentration for the subsequent enzymatic reactions.
-
[[File:eq3_tmao.png]]
+
[[File:eq3_tmao.png|300px|thumb|center]]
Regarding the enzymatic reaction:
Regarding the enzymatic reaction:
-
[[File:eq4_tmao.png]]
+
[[File:eq4_tmao.png|300px|thumb|center]]
-
The following relation holds E0 = Ce + C. And as usual imposing the steady state condition:
+
The following relation holds <html>E<sub>0</sub> = C<sub>E</sub> + C</html>. And as usual imposing the steady state condition:
-
[[File:eq5_tmao.png]]
+
[[File:eq5_tmao.png|300px|thumb|center]]
-
With Ce = E0 - C
+
With <html> C<sub>E</sub> = E<sub>0</sub> - C</html>
Solving for C:
Solving for C:
-
[[File:eq6_tmao.png]] with  
+
[[File:eq6_tmao.png|300px|thumb|center]] with [[File:eq7_tmao.png|300px|thumb|center|]]
 +
 
 +
In conclusion, we find:
 +
 
 +
[[File:eq8_tmao.png|300px|thumb|center|]]
 +
 
 +
 
 +
[[File:tmao_firstModeling.png|400px|thumb|center|]]
 +
 
 +
==Second Modeling==
 +
 
 +
It was used the same pattern of reactions as before, but with a standpoint of "essential activation" (the first reaction is no longer considered to equilibrium). Some enzymes need to be activated to bind the substrate. In our case, the enzyme is activated by TorS and activators are TMAO and TorT. The rate of product P appearance depends on the activators concentration.
 +
The equation that varies with respect to the previous case is:
 +
 
 +
[[File:eq9a_tmao.png|400px|thumb|center]]
 +
 
 +
 
 +
The total concentration of the enzyme is:
 +
 
 +
[[File:eq9_tmao.png|300px|thumb|center]]
 +
 
 +
Then we considered  [[File:eq10_tmao.png]]  and imposed [[File:eq11_tmao.png]] and [[File:eq12_tmao.png]]
 +
 
 +
The system of equations also comes from:
 +
 
 +
[[File:eq13_tmao.png|300px|thumb|center]]
 +
 
 +
Defining [[File:eq14_tmao.png]] and replacing we have:
 +
 
 +
[[File:eq15_tmao.png|300px|thumb|center]]
 +
 
 +
In conclusion the rate of product appearance results
 +
 
 +
[[File:eq16_tmao.png|300px|thumb|center]]
 +
 
 +
We have made simulations in order to see if our model can distinguish between different concentrations of TMAO (input) and how would be the activation of the transcription factor TorR (output). In general, it seems that the system will reach a saturation limit, except for  very low concentrations of TMAO, but the appearing of the product is faster for higher concentrations of TMAO.
 +
 
 +
[[File:tmao_simulations.jpeg|600px|thumb|center]]
 +
 
 +
=Modeling IMA System=
 +
 
 +
==First Reaction - Cobalt ions bind to albumin==
 +
 
 +
The cobalt ion binds to albumin protein, originating the cobalt-albumin complex. If the albumin is derived from a healthy patient, it will bind more cobalt ions than a albumin derived from a patient with heart disease. It happens because patients with heart conditions will have a isquemic modified albumin (IMA), which has less capability of binding to cobalt. Thus, the amount of free cobalt in the system will be determined by the presence of a normal or ischemic albumin in the sample.
 +
 
 +
[[File:eq1_cobalt.png|400px|thumb|center]]
 +
 
 +
==Second Reaction - The entry of cobalt in the cell==
 +
 
 +
There are four cobalt transporters in the cell. The rcnA (Uniprot code: P76425) and zntA (Uniprot code: P37617) transport the cobalt to outside the cell. The corA (Uniprot code: P0ABI4) and zupT (Uniprot code: P0A8H3) transport the cobalt to inside the cell.
 +
 
 +
[[File:eq3_cobalt.png|500px|thumb|center]]
 +
 
 +
The equation for the cobalt flux through these transporters is:
 +
 
 +
[[File:eq2_cobalt.png|500px|thumb|center]]
 +
 
 +
==Third Reaction - Transcription==
 +
 
 +
In our modeling, we considered the cobalt acting as a transcriptional activator in order to simplify our model.
 +
 
 +
The transcriptional activation are accessed by the equation:
 +
 
 +
[[File:eq4_cobalt.png|300px|thumb|center]]
 +
 
 +
The transcription and translation are accesed by the equations:
 +
 
 +
[[File:eq5_cobalt.png|300px|thumb|center]]
 +
 
 +
==Model simulations==
 +
 
 +
[[File:Cobalt_2_albumin_0_1.png|700px|thumb|center|Figure - Modeling simulation of IMA system]]
 +
 
 +
 
 +
The first simulation of our model showed that the outside concentration of cobalt reaches an equilibrium with the inside cobalt concentration of the bacterial cell. After a while, the Yellow Fluorescent Protein starts to appear. After that, we wanted to know how the system would behave with different concentrations of cobalt.
 +
 
 +
 
 +
[[File:Cobalt_concentrations_test.png|700px|thumb|center|Figure - Modeling simulation of IMA system with different concentrations of cobalt and 90 uM of Albumin protein ]]
 +
 
 +
 
 +
Simulating this situation it seems that YFP origin velocity reaches a limit for the higher concentrations of cobalt, as we see that there is just a little difference between the curves of 100, 125 e 150 uM of cobalt concentration. Thus we tested whether our model would be capable of distinguishing situations of higher and lower concentrations of normal albumin, as existent in healthy and affected individuals, respectively, according their heart condition.
 +
 
 +
 
 +
[[File:Cobalt_albumin_test.png|700px|thumb|center|Figure - Modeling simulation of IMA system with different concentrations of Albumin ]]
 +
 
 +
 
 +
The model could differentiate between the different concentrations of albumin, as shown in the graph above.
 +
 
 +
However, the pattern of our experimental results were different from the pattern obtained in our modeling. So, further caracterization and evaluation of our model is needed in order to obtain a better representation of the reality of our system.
-
[[File:eq7_tmao.png]].
+
==Model optimization==
-
In conclusion, we find
+
We have adapted an Evolutionary Algorithm to seek for the constants that
 +
best fit our model to the experimental results.
-
[[File:eq8_tmao.png]]
+
When we have functions that have an nonlinear behaviour,
 +
seeking for global optimal value, can be challenging and many types of functions and mathematical models
 +
have no known methods for this purpose.
 +
In these cases, it is fairly common to use stochastic methods, like Evolutionary Algorithms, to
 +
find interesting results within functions and models.
 +
An [http://en.wikipedia.org/wiki/Evolutionary_algorithm Evolutionary Algorithm] is a method in which each
 +
solution represents an individual and by crossing and mutating them, we can reach solutions that better fit
 +
our purpose.
 +
An schema, about how the Evolutionary Algorithm used our optimization works, is presented below:
 +
[[File:Full_ga_schema.png|700px|thumb|center|Figure - Evolutionary Algorithm schema [http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6557822&isnumber=6557545 Coelho, et al, 2013] ]]
 +
To fit the constants to the modeling, we compared the curve in our modeling to the
 +
curve obtained from experimental data and tried to reduce the difference between the curves.
 +
Despite the fact that the algorithm were able to reduce the difference between the curves,
 +
it couldn't find the curve we needed, because the modeling of YFP behaviour is not representing
 +
the experimental data behaviour.
{{Team:UFMG Brazil/sponsor}}
{{Team:UFMG Brazil/sponsor}}

Latest revision as of 13:31, 28 October 2013

Contents

Modeling TMAO system

Labels and conventions

CE: The enzymatic complex formed by TorT+TorS+TMAO, which reacts with TorR

C: intermediate component of the final enzymatic process

P: Product of the enzymatic reaction

CTS: Complex formed by TorT and TorS

CTSM: Complex formed by TorT+TorS and one molecule of TMAO

CTSM: Complex formed by TorT and TMAO

First Modeling

This is the simplest modeling process to be considered in our study. The product of the enzymatic process is collapsed in only one reaction that was placed at equilibrium.

Eq1 tmao.png

Imposing the equilibrium in the first reaction we have:

Eq2 tmao.png

The value of CE is the initial enzyme concentration for the subsequent enzymatic reactions.

Eq3 tmao.png

Regarding the enzymatic reaction:

Eq4 tmao.png

The following relation holds E0 = CE + C. And as usual imposing the steady state condition:

Eq5 tmao.png

With CE = E0 - C

Solving for C:

Eq6 tmao.png
with
Eq7 tmao.png

In conclusion, we find:

Eq8 tmao.png


Tmao firstModeling.png

Second Modeling

It was used the same pattern of reactions as before, but with a standpoint of "essential activation" (the first reaction is no longer considered to equilibrium). Some enzymes need to be activated to bind the substrate. In our case, the enzyme is activated by TorS and activators are TMAO and TorT. The rate of product P appearance depends on the activators concentration. The equation that varies with respect to the previous case is:

Eq9a tmao.png


The total concentration of the enzyme is:

Eq9 tmao.png

Then we considered Eq10 tmao.png and imposed Eq11 tmao.png and Eq12 tmao.png

The system of equations also comes from:

Eq13 tmao.png

Defining Eq14 tmao.png and replacing we have:

Eq15 tmao.png

In conclusion the rate of product appearance results

Eq16 tmao.png

We have made simulations in order to see if our model can distinguish between different concentrations of TMAO (input) and how would be the activation of the transcription factor TorR (output). In general, it seems that the system will reach a saturation limit, except for very low concentrations of TMAO, but the appearing of the product is faster for higher concentrations of TMAO.

Tmao simulations.jpeg

Modeling IMA System

First Reaction - Cobalt ions bind to albumin

The cobalt ion binds to albumin protein, originating the cobalt-albumin complex. If the albumin is derived from a healthy patient, it will bind more cobalt ions than a albumin derived from a patient with heart disease. It happens because patients with heart conditions will have a isquemic modified albumin (IMA), which has less capability of binding to cobalt. Thus, the amount of free cobalt in the system will be determined by the presence of a normal or ischemic albumin in the sample.

Eq1 cobalt.png

Second Reaction - The entry of cobalt in the cell

There are four cobalt transporters in the cell. The rcnA (Uniprot code: P76425) and zntA (Uniprot code: P37617) transport the cobalt to outside the cell. The corA (Uniprot code: P0ABI4) and zupT (Uniprot code: P0A8H3) transport the cobalt to inside the cell.

Eq3 cobalt.png

The equation for the cobalt flux through these transporters is:

Eq2 cobalt.png

Third Reaction - Transcription

In our modeling, we considered the cobalt acting as a transcriptional activator in order to simplify our model.

The transcriptional activation are accessed by the equation:

Eq4 cobalt.png

The transcription and translation are accesed by the equations:

Eq5 cobalt.png

Model simulations

Figure - Modeling simulation of IMA system


The first simulation of our model showed that the outside concentration of cobalt reaches an equilibrium with the inside cobalt concentration of the bacterial cell. After a while, the Yellow Fluorescent Protein starts to appear. After that, we wanted to know how the system would behave with different concentrations of cobalt.


Figure - Modeling simulation of IMA system with different concentrations of cobalt and 90 uM of Albumin protein


Simulating this situation it seems that YFP origin velocity reaches a limit for the higher concentrations of cobalt, as we see that there is just a little difference between the curves of 100, 125 e 150 uM of cobalt concentration. Thus we tested whether our model would be capable of distinguishing situations of higher and lower concentrations of normal albumin, as existent in healthy and affected individuals, respectively, according their heart condition.


Figure - Modeling simulation of IMA system with different concentrations of Albumin


The model could differentiate between the different concentrations of albumin, as shown in the graph above.

However, the pattern of our experimental results were different from the pattern obtained in our modeling. So, further caracterization and evaluation of our model is needed in order to obtain a better representation of the reality of our system.

Model optimization

We have adapted an Evolutionary Algorithm to seek for the constants that best fit our model to the experimental results.

When we have functions that have an nonlinear behaviour, seeking for global optimal value, can be challenging and many types of functions and mathematical models have no known methods for this purpose. In these cases, it is fairly common to use stochastic methods, like Evolutionary Algorithms, to find interesting results within functions and models.

An [http://en.wikipedia.org/wiki/Evolutionary_algorithm Evolutionary Algorithm] is a method in which each solution represents an individual and by crossing and mutating them, we can reach solutions that better fit our purpose.

An schema, about how the Evolutionary Algorithm used our optimization works, is presented below:

Figure - Evolutionary Algorithm schema [http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6557822&isnumber=6557545 Coelho, et al, 2013]

To fit the constants to the modeling, we compared the curve in our modeling to the curve obtained from experimental data and tried to reduce the difference between the curves.

Despite the fact that the algorithm were able to reduce the difference between the curves, it couldn't find the curve we needed, because the modeling of YFP behaviour is not representing the experimental data behaviour.

Our Sponsors

Reitoria-de-pesquisa-UFMG.jpg Reitoria-de-posgraduacao-UFMG.jpg Icb ufmg.jpg Bioquimica.jpg Bioinformatica.jpg INCT.jpg Inctv.jpg Nanobiofar.jpg Fapemig.jpg Sintesebiotecnologia.jpg