Difference between revisions of "Team:Manchester/Model"

 
(45 intermediate revisions by 3 users not shown)
Line 8: Line 8:
 
     width:90%;
 
     width:90%;
 
     margin:auto;
 
     margin:auto;
−
}
 
−
 
−
area:hover{
 
−
    outline: 1px blue solid;
 
 
}
 
}
  
Line 90: Line 86:
 
  </div>
 
  </div>
  
−
<center>
 
−
  <div class="column full_size frontword">
 
−
    <p style="font-size:1.2em;">
 
−
Welcome to our modelling section. We have used a novel ensemble modeling approach, to better aid the synergy between wetalb and dry lab teams. On this page you will find short answers to the questions; What is ensemble modelling? What did we model? what did our model achieve?
 
−
      <br /><br />
 
−
For navigating the Wiki you need to know that the sections on results and human practice/lab  integration can be accessed in the menu bar.
 
−
    <br /><br />
 
−
Part of what we hope to achieve with this ensemble methodology is a blueprint for other iGEM teams. As such each step in creating and using our model is laid out in the below diagram. Clicking on a specific step will take you to a page explaining the theory. Going deeper you can access a discussion of the code. There will be a link to the relevant parts of our github (a website storing our code) Use these codes as you wish.
 
−
    <br /><br />
 
−
<a href="https://github.com/Manchester-iGem-2016/UoMiGem2016">Link to github homepage</a> 
 
  
−
    </p>
 
  
 +
<center>
 +
<div class="width90">
  
−
</div>
 
  
−
</center>
 
  
  
 +
<div class="column onethird_size">
 +
<a href="#model1"><img class="model1" src="https://static.igem.org/mediawiki/2016/c/c1/T--Manchester--modelling1_pic.png" alt="What were we modelling?"></img></a>
 +
</div>
  
 +
<div class="column onethird_size" >
 +
<a href="#model2"><img class="model1" src="https://static.igem.org/mediawiki/2016/a/a3/T--Manchester--modelling2_pic.png" alt="What is esemble modelling?"></img></a>
 +
</div>
  
  
−
<!--
+
<div class="column onethird_size" >
−
<div class="column twothird_size">
+
<a href="#model3"><img class="model1" src="https://static.igem.org/mediawiki/2016/4/46/T--Manchester--modelling3_pic.png" alt="What does our modelling achieve?"></img></a>
−
<h5 class="modelling_info2">
+
</br></br>
−
We focused on modelling the <a href="https://2016.igem.org/Team:Manchester/Description/mechanism1">Cell-free Mechanism</a>, see the experimental section for details. The short version is the AlcoPatch relies on alcohol, alcohol oxidase (AOx), horseradish peroxidase (HRP) and 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) (ABTS) which are mixed together making oxidised ABTS which is colourful. The ensemble modelling process will work for any model however and this was chosen as a simpler example to demonstrate the technique, focusing on the process rather than the system itself.
+
−
 
+
−
<br /><br />
+
−
All the modelling and <a href="https://2016.igem.org/Team:Manchester/Description">experimental work</a> for the <a href="https://2016.igem.org/Team:Manchester/Description/mechanism1">Cell-free Mechanism</a> was done using glucose and glucose oxidase (GOx) instead, this was due to laboratory limitations. This effects the modelling analysis in trying to predict things about alcohol which came from human practices. The equivalent analysis was done for glucose. Rerunning the analysis for alcohol would only require the change of some constants, so the analysis acts as a proof of concept and still shows the integration of human practices.
+
−
 
+
−
</h5>
+
 
</div>
 
</div>
−
-->
 
  
  
  
−
<!--
+
<div id="model3">
−
<div class="column twothird_size">
+
−
<h5 class="modelling_info1">
+
−
Instead of running your model once with some specific parameters (i.e rate constants), you find all the possible parameters in literature and run your model for lots of combinations sampled in a clever way. By doing this you take take into account the uncertainty in the parameters. You will create probabilistic outputs allowing you to make rigorous conclusions.
+
−
            <br /><br />
+
−
Following through our flow diagram blueprint should make this concept clear alternatively click the steps from the list below..
+
−
</h5>
+
 
</div>
 
</div>
−
-->
 
  
 +
<div  class="modelling_info1">
 +
<p style="border-bottom: 1px black solid ;font-size:25px;text-weight:bold;display:inline-block">What did our model achieve?</p>
 +
<p style="font-size:1.2em;">
 +
We achieved 3 main aims in our modelling work:
  
  
 +
</br> We introduced a novel <a href="https://2016.igem.org/Team:Manchester/Model/ModelExplorer">ensemble modelling</a> approach to iGEM and made this approach accessible to other iGEM teams by sharing <a target="_Blank" href="https://github.com/Manchester-iGem-2016/Ensemble-Modelling">our code.</a>
 +
<br />
 +
We improved our understanding of our system and used real experimental data to improve our model, using <a href="https://2016.igem.org/Team:Manchester/Model/MechanismUncertainty">network mechanism analysis </a> and <a href="https://2016.igem.org/Team:Manchester/Model/ParameterRelationships">parameter relationship analysis</a>. <br />
 +
We answered key questions that arose during our <a href="https://2016.igem.org/Team:Manchester/Model/hp"> integrated human practices</a> work, helping to improve the design of our system using <a href="https://2016.igem.org/Team:Manchester/Model/Costing">cost analysis</a>.
  
  
−
<!--
+
</br></br>
−
<div class="column twothird_size">
+
All of our models are available on <a target="_Blank" href="https://github.com/Manchester-iGem-2016/Ensemble-Modelling">our Github page</a>
−
<h5 class="modelling_info3">
+
−
Firstly we narrowed down the potential reaction schemes and increased our knowledge of our rate constants.
+
−
Secondly we have created a blueprint for using ensemble modelling instead of simple ODE modelling in the hope that it can become standard practice for iGEM. <!--MATT THIS WAS A PLACE HOLDER YOU ARE TO WRITE THIS AFTER THE RESULTS SECTION AS YOU WILL BE ABLE TO LINK BETWEEN THE TOO. -->    
+
−
</h5>
+
−
</div>
+
−
-->
+
−
 
+
−
 
+
  
 +
</p>
  
 +
<table>
 +
      <th><a href="https://2016.igem.org/Team:Manchester/Model/MechanismUncertainty">Network Mechanism Analysis </a></th>
 +
      <th><a href="https://2016.igem.org/Team:Manchester/Model/ParameterRelationships">Parameter Relationship Analysis</a></th>
 +
      <th><a href="https://2016.igem.org/Team:Manchester/Model/Costing">Cost Analysis</a></th>
 +
<tr>
 +
    <td>Comparing model predictions with experimental data for different potential circuit topologies. <a href="https://2016.igem.org/Team:Manchester/Model/MechanismUncertainty">read more</a></td>
 +
    <td>Assessing the interlinking nature of specific parameter pairings on the outcomes of the system. <a href="https://2016.igem.org/Team:Manchester/Model/ParameterRelationships"> read more </a></td>
 +
    <td>Predicting the costs for a range of different system specifications by varying the amount of enzymes based on experimental data. <a href="https://2016.igem.org/Team:Manchester/Model/Costing">read more</a></td>
 +
</tr>
 +
</table>
  
  
 +
<div class="modelling_info1">
 
<center>
 
<center>
−
<div class="width90">
+
<p style="font-size:1.2em;">
−
 
+
We found great inspiration from our <a href="https://2016.igem.org/Team:Manchester/Human_Practices/Industries">human practices</a> and guidance working both ways with the experiments. Click <a href="https://2016.igem.org/Team:Manchester/Model/hp"> here</a> to see a summary.
−
 
+
</p>
−
 
+
</center>
−
 
+
−
<div class="column onethird_size">
+
−
<a href="#model1"><img class="model1" src="https://static.igem.org/mediawiki/2016/c/c1/T--Manchester--modelling1_pic.png" alt="What were we modelling?"></img></a>
+
 
</div>
 
</div>
  
−
<div class="column onethird_size" >
 
−
<a href="#model2"><img class="model1" src="https://static.igem.org/mediawiki/2016/a/a3/T--Manchester--modelling2_pic.png" alt="What is esemble modelling?"></img></a>
 
 
</div>
 
</div>
−
 
−
 
−
<div  class="column onethird_size" >
 
−
<a href="#model3"><img class="model1" src="https://static.igem.org/mediawiki/2016/4/46/T--Manchester--modelling3_pic.png" alt="What does our modelling achieve?"></img></a>
 
−
 
 
</div>
 
</div>
  
Line 180: Line 159:
 
<p style="border-bottom: 1px black solid ;font-size:25px;text-weight:bold;display:inline-block">What were we modelling?</p>
 
<p style="border-bottom: 1px black solid ;font-size:25px;text-weight:bold;display:inline-block">What were we modelling?</p>
 
<p style="font-size:1.2em;">
 
<p style="font-size:1.2em;">
−
We focused on modelling the <a href="https://2016.igem.org/Team:Manchester/Description/mechanism1">Cell-free Mechanism</a>, see the <a href="https://2016.igem.org/Team:Manchester/Description/mechanism1">experimental</a> section for details. The short version is the AlcoPatch relies on alcohol, alcohol oxidase (AOx), horseradish peroxidase (HRP) and 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) (ABTS) which are mixed together producing oxidised ABTS which is colourful. The ensemble modelling process will work for any model however and this was chosen as a simpler example to demonstrate the technique, focusing on the process rather than the system itself.  
+
We focused on modelling the <a href="https://2016.igem.org/Team:Manchester/Description/mechanism1">Cell-free Mechanism</a>. The short version is the AlcoPatch relies on alcohol, alcohol oxidase (AOx), horseradish peroxidase (HRP) and 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) (ABTS) to detect and quantify alcohol levels. The ABTS<sub>Oxidised</sub> produced in the prescence of alcohol is dark green and can be detected spectophotometrically or visually.
 +
</br></br>
 +
We focused on this small system because it was possible to obtain a large amount of <a href="https://2016.igem.org/Team:Manchester/Notebook">experimental data</a> for model validation, and because it allowed us to establish and illustrate the <a href="https://2016.igem.org/Team:Manchester/Model/ModelExplorer">ensemble modelling</a> process.
 
</p>
 
</p>
 
<p style="font-size:1.2em;">
 
<p style="font-size:1.2em;">
−
Due to time constraints the modelling was based on an analogous system of glucose and glucose oxidase (GOx) rather than alcohol and AOx, this was chosen as the reaction network is the same and it links in with the suggestions given at the <a href="https://2016.igem.org/Team:Manchester/Human_Practices/Outreach#microbio">Microbiology Society Conference</a> in the <a href="https://2016.igem.org/Team:Manchester/Human_Practices">human practices</a> that the design need not be limited for alcohol but could be used by diabetics, etc if used to detect other molecules in sweat.
+
The majority of our experimental data came from the <a href="https://2016.igem.org/Team:Manchester/Proof">proof-of-concept</a> study of the analogous system of glucose and glucose oxidase (GOx) rather than alcohol and AOx. The reaction network of the two sytems is the same and only some kinetic parameters differ.
−
While the equivalent analysis was done for glucose. Rerunning the analysis for alcohol would only require the change of some constants, so the analysis acts as a sufficient proof of concept and still shows the <a href="https://2016.igem.org/Team:Manchester/Integrated_Practices">integration of human practices</a>.
+
 
</p>
 
</p>
 
<p style="font-size:1.2em;">
 
<p style="font-size:1.2em;">
−
A schematic diagram of the final scheme is given below. For more information about the steps click on the blue enzyme boxes.
+
A schematic diagram of the final circuit of our detection system is given below. </br>
 +
<b> For more information about the individual reactions click on the blue enzyme boxes. </b>
 
</p>
 
</p>
 
<img class="full" src="https://static.igem.org/mediawiki/2016/0/04/T--Manchester--ModellingNetworkDiagram.png" alt="Reaction Network Diagram used in the modelling" usemap="#diagramclick" />
 
<img class="full" src="https://static.igem.org/mediawiki/2016/0/04/T--Manchester--ModellingNetworkDiagram.png" alt="Reaction Network Diagram used in the modelling" usemap="#diagramclick" />
Line 214: Line 195:
 
<p style="border-bottom: 1px black solid ;font-size:25px;text-weight:bold;display:inline-block">What is Ensemble Modelling?</p>
 
<p style="border-bottom: 1px black solid ;font-size:25px;text-weight:bold;display:inline-block">What is Ensemble Modelling?</p>
 
<p style="font-size:1.2em;">
 
<p style="font-size:1.2em;">
−
We model the reactions in our AlcoPatch system using ordinary differential equations that is dependent on kinetic parameters and have used it to generate a composite of predictions using ensemble modelling. </br>
+
Incomplete and uncertain knowledge of kinetic parameters is a common problem when building models for synthetic biology. Ensemble modelling is one strategy to deal with this problem. Instead of running our model with a single set of specific parameters (for example rate constants), we run our model multiple times using different sets of plausible parameter values and analyse the predictions as an ensemble. We collected all the available parameter values from published literature and took into account the uncertainties that are associated with them. The resulting confidence in our parameter values was then described by <a href="https://2016.igem.org/Team:Manchester/Model/PDF">probability density functions</a>. </br>
  
−
As multiple and various kinetic values was found for the parameters, each parameter is described by a probability distribution which expresses the associated uncertainty. Sampling from these probability distributions have allowed our AlcoPatch model to simulate its reactions with different sets of plausible kinetic values. This means that instead of running a single prediction from a fixed single set of kinetic parameter values, we could run our AlcoPatch model a number of times to generate multiple predictions that we later analyse as an ensemble to provide probabilistic predictions of the system. </br>
+
This has created probabilistic outputs allowing us to make rigorous conclusions about our reaction mechanism – and to assess which predictions are reliable, and where we are lacking information. </br> </br>
  
−
Ensemble modelling has yet to ‘breakthrough’ into the world of biological modelling and systems biology in general. Essentially, we analyse multiple predictions from our model as a composite, which is different from traditional predictive modelling methods. </br>
+
<b>To explore the theory of this process please click the boxes on the diagram below.</b>
−
Instead of running a single prediction from a fixed single set of kinetic parameter values, we run our AlcoPatch model a number of times from different sets of kinetic values by sampling from the probability distributions generated from the previous step. We then analyse the entire set of predictions as an ensemble to draw probabilistic conclusions about the system. </br>
+
−
To explore the theory of this process please click the boxes on the diagram below.
+
  
 
</p>
 
</p>
Line 230: Line 209:
 
<map name="diagram1click">
 
<map name="diagram1click">
 
<area shape="rect" coords="70,82,413,190" href="https://2016.igem.org/Team:Manchester/Model/ParameterSelection" title="Parameter Selection">
 
<area shape="rect" coords="70,82,413,190" href="https://2016.igem.org/Team:Manchester/Model/ParameterSelection" title="Parameter Selection">
−
<area shape="rect" coords="505,85,858,188" href="https://2016.igem.org/Team:Manchester/Model/PDF" title="PDF">
+
<area shape="rect" coords="505,83,861,188" href="https://2016.igem.org/Team:Manchester/Model/PDF" title="PDF">
−
<area shape="rect" coords="630,280,860,380" href="https://2016.igem.org/Team:Manchester/Model/Simulate" title="system Simulate">
+
<area shape="rect" coords="635,305,850,411" href="https://2016.igem.org/Team:Manchester/Model/Simulate" title="system Simulate">
−
<area shape="rect" coords="360,300,580,400" href="https://2016.igem.org/Team:Manchester/Model/Analyse" title="Result Analysis">
+
<area shape="rect" coords="338,307,553,412" href="https://2016.igem.org/Team:Manchester/Model/result" title="Result Analysis">
−
<area shape="rect" coords="50,298,265,400" href="https://2016.igem.org/Team:Manchester/Model/Update" title="Update Model">
+
<area shape="rect" coords="50,298,265,400" href="https://2016.igem.org/Team:Manchester/Model/Story" title="Update Model">
 
</map>
 
</map>
  
Line 246: Line 225:
 
<a href="https://2016.igem.org/Team:Manchester/Model/PDF">Generating Probability Density Functions</a> </br>
 
<a href="https://2016.igem.org/Team:Manchester/Model/PDF">Generating Probability Density Functions</a> </br>
 
<a href="https://2016.igem.org/Team:Manchester/Model/Simulate">Simulate the System</a> </br>
 
<a href="https://2016.igem.org/Team:Manchester/Model/Simulate">Simulate the System</a> </br>
−
<a href="https://2016.igem.org/Team:Manchester/Model/Analyse">Analyse the Results</a> </br>
+
<a href="https://2016.igem.org/Team:Manchester/Model/result">Analyse the Results</a> </br>
−
<a href="https://2016.igem.org/Team:Manchester/Model/Update">Update the Model</a> </br>
+
<a href="https://2016.igem.org/Team:Manchester/Model/Story">Story of the Model</a> </br>
 
</p>
 
</p>
 
</div>
 
</div>
 +
</centre>
  
 +
<span class="box">
  
 +
</span>
  
−
<div id="model3">
+
</body>
−
</div>
+
−
 
+
−
<div  class="modelling_info1">
+
−
<p style="border-bottom: 1px black solid ;font-size:25px;text-weight:bold;display:inline-block">What did our model achieve?</p>
+
−
<p style="font-size:1.2em;">
+
−
Regardless how clever, interesting or unique a model is ultimately all that matters is the results at the end. We had 2 main aims for our outputs: improving our understanding of our system and answering key questions that arouse during the Human Practices.
+
−
</p>
+
−
<p style="font-size:1.2em;">
+
−
To improve our knowledge of the system we undertook 2 main analyses: </br>
+
−
<a href="https://2016.igem.org/Team:Manchester/Model/MechanismUncertainty">Improving our understanding the reaction network mechanism</a> </br>
+
−
<a href="https://2016.igem.org/Team:Manchester/Model/ParameterRelationships">Investigating the relationships between the parameters in our system</a> </br> </br>
+
−
To answer the questions from the <a href="https://2016.igem.org/Team:Manchester/Human_Practices/Industries">human practices</a> there was 1 main analysis: </br>
+
−
<a href="https://2016.igem.org/Team:Manchester/Model/Costing">Costing the AlcoPatch</a> </br></br></br>
+
−
Further justification for why these were chosen is given on the respective pages.
+
−
</p>
+
−
</div>
+
−
 
+
−
 
+
−
 
+
−
 
+
−
 
+
−
<div class="modelling_info1">
+
−
<center>
+
−
<p style="font-size:1.2em;">
+
−
We found great inspiration from our <a href="https://2016.igem.org/Team:Manchester/Human_Practices/Industries">human practices</a> and guidance working both ways with the experiments. Click <a href="https://2016.igem.org/Team:Manchester/Model/hp"> here</a> to see a summary.
+
−
</p>
+
 
</center>
 
</center>
−
</div>
 
  
 +
<div class="floatleft1 project1">
 +
<a class="projectlink" href="https://2016.igem.org/Team:Manchester"><< Main Page</a>
 
</div>
 
</div>
−
</center>
 
−
</div>
 
−
 
−
 
−
<span class="box1">
 
−
 
−
</span>
 
−
 
−
</body>
 
  
 
<script>
 
<script>

Latest revision as of 14:50, 24 November 2016

Manchester iGEM 2016
Modelling Banner
What were we modelling?
What is esemble modelling?
What does our modelling achieve?

What did our model achieve?

We achieved 3 main aims in our modelling work:
We introduced a novel ensemble modelling approach to iGEM and made this approach accessible to other iGEM teams by sharing our code.
We improved our understanding of our system and used real experimental data to improve our model, using network mechanism analysis and parameter relationship analysis.
We answered key questions that arose during our integrated human practices work, helping to improve the design of our system using cost analysis.

All of our models are available on our Github page

Network Mechanism Analysis Parameter Relationship Analysis Cost Analysis
Comparing model predictions with experimental data for different potential circuit topologies. read more Assessing the interlinking nature of specific parameter pairings on the outcomes of the system. read more Predicting the costs for a range of different system specifications by varying the amount of enzymes based on experimental data. read more

We found great inspiration from our human practices and guidance working both ways with the experiments. Click here to see a summary.

What were we modelling?

We focused on modelling the Cell-free Mechanism. The short version is the AlcoPatch relies on alcohol, alcohol oxidase (AOx), horseradish peroxidase (HRP) and 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) (ABTS) to detect and quantify alcohol levels. The ABTSOxidised produced in the prescence of alcohol is dark green and can be detected spectophotometrically or visually.

We focused on this small system because it was possible to obtain a large amount of experimental data for model validation, and because it allowed us to establish and illustrate the ensemble modelling process.

The majority of our experimental data came from the proof-of-concept study of the analogous system of glucose and glucose oxidase (GOx) rather than alcohol and AOx. The reaction network of the two sytems is the same and only some kinetic parameters differ.

A schematic diagram of the final circuit of our detection system is given below.
For more information about the individual reactions click on the blue enzyme boxes.

Reaction Network Diagram used in the modelling

Alternatively you can click on the enzyme name below:

Glucose Oxidase
Horseradish Peroxidase

What is Ensemble Modelling?

Incomplete and uncertain knowledge of kinetic parameters is a common problem when building models for synthetic biology. Ensemble modelling is one strategy to deal with this problem. Instead of running our model with a single set of specific parameters (for example rate constants), we run our model multiple times using different sets of plausible parameter values and analyse the predictions as an ensemble. We collected all the available parameter values from published literature and took into account the uncertainties that are associated with them. The resulting confidence in our parameter values was then described by probability density functions.
This has created probabilistic outputs allowing us to make rigorous conclusions about our reaction mechanism – and to assess which predictions are reliable, and where we are lacking information.

To explore the theory of this process please click the boxes on the diagram below.

Overview flowchart of ensemble modelling

Alternatively you can click on the step name below:

Collecting and Processing Data
Generating Probability Density Functions
Simulate the System
Analyse the Results
Story of the Model

<< Main Page

Sponsors

Follow us on