Best of 2006-2007
Best of 2007-2008
What are the biological analogues of Sensors and Actuators?
- let's leave it pretty open: Maybe you identify a biological system where the sensor and actuator are distinctly separate. Is the system modular? (Could you replaced one sensor with another?, likewise w/ actuators) Or maybe an example of a system where this division is blurred, and possibly explain why it's useful. Any way you choose to interpret it, continue to bring interesting examples:)
- Nathan volunteered to present this paper on Eukaryotic Chemotaxis:
- The van Oudenaarden paper is another nice treatment of eukaryotic chemotaxis (particularly dealing with the question of noise in such systems).-Jkm
Error fetching PMID 15691763:
Error fetching PMID 16864788:
- Error fetching PMID 15691763:
- Error fetching PMID 16864788:
- post interesting examples here before Friday...
- RMM: here are some random thoughts that I'll update during the week
- Sensors: this seems pretty straightforward - there are lots of examples of molecules that bind to other molecules and "sense" the presence of those molecules. However the "signal" out of the sensor is less obvious. Sometimes you get a conformation change that enables the molecule to do something new and different. Sometimes you block an event from occuring and that is the "signal" corresponding to that sensor.
- Actuators: I guess the big one in cells is somehow affecting the transcription or translation process.
- NWP: Thought experiments to ponder for Friday and while reading paper -
- If you were engineering chemotaxis where would you apply gain to the existing system to improve performance?
- Simple sounding task, How would you engineer a cell to move away from a chemoattranctant instead of towards?
- How does time delay affect the stability of the feedback system and the choice of gain placement in this system?
- If we made an analogy to PID controllers, are there combinations we can rule out and why?
- High robustness comes with a heavy price, i.e. fragility somewhere in the system… How does biased stochasticity create dynamic instability in this system? Or put another way, how does the cell use fragility to its advantage? Fragility to robustness conversion, eh?
- What are the equilibrium points for this system? Or put another way, does the system show any stable states which if the cell comes close to, it stays close? And do these equilibrium points depend on any of the system parameters?
- Can anyone think of engineering systems that use this type of control?
- Josh: Some comments:
- Sensing (particularly gradient sensing in chemotaxis) is more complicated than it might seem. Prokaryotes, for instance, are too small to sense spatial gradients, so they have to sense temporal gradients instead.
- Actuation can happen much faster by affecting protein function directly (allostery) - think kinase/phosphatases.
- Eukaryotic - make the probability of lamellipod extension decrease for increasing chemorepellent.
- Prokaryotic - make the tumble probability increase for increasing chemorepellent.
- Gain: Noise should quickly become a problem - both stochastics (the paper seems to have simulated ~3 lamellipodia per cell at any given time) and environmental fluctuations (need to damp out fast fluctuations, or else you just spin around in place).
- Robustness: I'd say you trade off robustness and performance, rather than fragility (not sure what exactly you mean by fragility, incidentally). If the cell picked one direction and went that way as fast as it could, it'd get there faster, but some times it would head in the wrong direction. It's interesting that both prokaryotic and eukaryotic chemotaxis convergently evolved to a similar biased random walk. Suggests to me that this is pretty close to an optimal solution to the physical problem.