Week8 : Test execution of simulation for large model : red blood cell model

Hi all

1. What I did this week
I mainly focused on the test of simulation execution for large model: red blood cell model[1]. I tried to execute simulation.

1.1 Demo of simulation for red blood cell model

Fig1.1.1 : Demo of simulation for red blood cell model

As you can see, the dynamic range of the expression for these species are wide. Therefore, the result is really difficult to see. So I tried to set the form to change it logarithmic scale.

Fig1.1.2 : The logarithmic scale to visualize the simulation for red blood cell model.

This implementation is also difficult to see. I will discuss it in [Discussion]. Moreover, I faced the other problem to scale logarithmic

Fig1.1.3 : If the end time is changed to more than 60 for red blood cell model, the logarithmic scaling is crushed.

This is specifically happened in the condition following Fig.1.1.3. I have already confirmed that logarithmic scaling is correctly worked for other models in almost all of simulation condition.

Fig1.1.4 : The logarithmic scaling to visualize the simulation for MAPK model is correctly worked. I also confirmed the logarithmic scaling for caffein model and sample model in example directory in my repository is correctly worked.

[Discussion]
1) The crush of logarithmic scaling
I used Chart.js to visualize the simulation result and I only changed the type to "logarithmic" in yAxes which is object of Chart.js. Unfortunately I have no idea to solve it. In meeting, I will discuss it with my mentors.

2) Even if the result is scaled logarithmically, the simulation result is difficult to see.
As you can see Fig1.1.2, this is totally difficult to see the result of simulation. An idea to solve it is that the range of concentration is limited to the species which user chooses. The UI to select the species whose dynamical change of concentration is visualized have already been implemented. Therefore, I can easily implement it.

2. The task of next week
I planned following tasks.

1) Implementation of the numerical visualization of simulation and parameter estimation
2) Fixing the problem of crush of logarithmic scaling

Thanks
Reference:
[1] Bordbar, Aarash, et al. "Personalized whole-cell kinetic models of metabolism for discovery in genomics and pharmacodynamics." Cell systems 1.4 (2015): 283-292.

コメント

このブログの人気の投稿

Final Week : Create document and implement some minor UI

Week9 : Numerical table of the result from simulation and parameter estimation

Week12 : Download widget was created and all javascript code to execute ajax transfer is changed based on JQuery