ETHZ/Model

From 2007.igem.org

(Difference between revisions)
Line 1: Line 1:
-
For our poject we decided on designing a systems that is able to learn or adapt to its environment. Please note that this is only a minimal system that should be able to act as a proof of concept.
+
For our poject we decided on designing a systems that is able to learn or adapt to its environment. Please note that this is only a minimal system that should be able to act as a proof of concept. A protocol how the system should react according to an input is shown in Figure 1.
[[Image:ETHZFlowdiagram.png|center|thumb|Fig. 1: Flow diagram |480px]]
[[Image:ETHZFlowdiagram.png|center|thumb|Fig. 1: Flow diagram |480px]]
 +
 +
The idea behind this protocol is that
 +
* The system will be able to learn one of two input signals - aTc or IPTG - furing a learning phase if no input signal AHL is present. Depending on the input it will report by either green or yellow florescence.
 +
* Once the system learned, the inputs - aTc or IPTG - can be released and the system goes into a memory state in the presence of AHL. In this state no output color is reported.
 +
* During a recognition phase the inputs aTc or IPTG are (re-) presented again. The system reports by changing its color depending on the input and its current memory state. That is why the system can have different florescence properties even in the presence of the same input. The recognition phase takes place in the presence of AHL to keep the memory enabled and avoid another learning phase.
 +
 +
 +
To define our system we start with the classical back box approach as shown in Figure 2.
[[Image:ETHZBlackbox.png|center|thumb|Fig. 2: Black box |480px]]
[[Image:ETHZBlackbox.png|center|thumb|Fig. 2: Black box |480px]]
-
proof of concept for a minmal learning/adapting system
+
To fill that back box we have to think a bit more about the properties of our system. From our protocol we know that we need:
 +
* 2 inputs that should be learned/detected/adapted to,
 +
* 1 input to switch on the memory.
 +
* We need to store at least 3 states. That is why we decided to use 2 state variables.
 +
* To be more robust against pertubations we couple that state variables in the way that is well known from memory circuits that engineers build where one  state variable is depressing the other one.
 +
* We need 4 florescense signals for the outputs. Actually one could also decide to take 6 output signals into account to further distinguish the learning phase from the recognition phase but we restricted ourself to 4 outputs to reduce the number of genes that we will need to implement the signals.
-
what do we need?
+
An overview about the final system is shown in Figure 3.
-
* 2 inputs that should be learned/detected/adapted to
+
-
* 1 input to switch on memory
+
-
* need to store 3 states -> 2 state variables
+
-
* couple state variables to be more robust against pertubations
+
-
* output 4 florescense proteins
+
[[Image:ETHZFullsystemmodel.png|center|thumb|Fig.3: System overview |480px]]
[[Image:ETHZFullsystemmodel.png|center|thumb|Fig.3: System overview |480px]]

Revision as of 09:42, 15 October 2007

For our poject we decided on designing a systems that is able to learn or adapt to its environment. Please note that this is only a minimal system that should be able to act as a proof of concept. A protocol how the system should react according to an input is shown in Figure 1.

Fig. 1: Flow diagram

The idea behind this protocol is that

  • The system will be able to learn one of two input signals - aTc or IPTG - furing a learning phase if no input signal AHL is present. Depending on the input it will report by either green or yellow florescence.
  • Once the system learned, the inputs - aTc or IPTG - can be released and the system goes into a memory state in the presence of AHL. In this state no output color is reported.
  • During a recognition phase the inputs aTc or IPTG are (re-) presented again. The system reports by changing its color depending on the input and its current memory state. That is why the system can have different florescence properties even in the presence of the same input. The recognition phase takes place in the presence of AHL to keep the memory enabled and avoid another learning phase.


To define our system we start with the classical back box approach as shown in Figure 2.

Fig. 2: Black box

To fill that back box we have to think a bit more about the properties of our system. From our protocol we know that we need:

  • 2 inputs that should be learned/detected/adapted to,
  • 1 input to switch on the memory.
  • We need to store at least 3 states. That is why we decided to use 2 state variables.
  • To be more robust against pertubations we couple that state variables in the way that is well known from memory circuits that engineers build where one state variable is depressing the other one.
  • We need 4 florescense signals for the outputs. Actually one could also decide to take 6 output signals into account to further distinguish the learning phase from the recognition phase but we restricted ourself to 4 outputs to reduce the number of genes that we will need to implement the signals.

An overview about the final system is shown in Figure 3.

Fig.3: System overview
Retrieved from "http://2007.igem.org/wiki/index.php/ETHZ/Model"