# User:J. C. Martinez-Garcia/Notebook/HMS Activities/2008/11/03

### From OpenWetWare

Coming back to the raw data | Main project page Previous entry Next entry |

## Working from the scratchI decided to work in the modeling task concerning the modified AKT signaling following a raw data approach. I asked to David to give the original data that he obtained from the experiments that he performed. He send me a message with the requested data, which are given in the next figure: Then, I decided to follow three different approaches: - Design a neural network just to match the data.
- Model the modified AKT signaling pathway using the SimBiology toolbox of MatLab.
- Directly obtain a model using the nonlinear differential equations formalism.
I think that a perceptron based network will be good for the first purpose. I suppose that a network can be configured in order to have two different inputs (the Zstk drug concentration and the mTOR raptor inhibitor). The idea is to have a black-box model which display the two different given behaviors (with or without mTOR raptor inhibitor). For the second approach the advantage would be to have a model easily available for anybody working with the Matlab framework. ## Meeting with PilpelTomorrow I will read again the Pilpel papers. I asked him to have a meeting next Wednesday. |