Evolution of efficiency

Sir-Bip-Bop / Gap-Junction-Coupling-Studies

Repository with the code of the masters project and a link to the overleaf document, to be able to move through both quickly

0 stars 0 forks source link

Evolution of efficiency #3

Open Sir-Bip-Bop opened 1 year ago

Sir-Bip-Bop commented 1 year ago

izhikevich has a higher value for gap junction strength, and morris-lecar a lower one. Coincidentally, their efficiencies are also higher and lower respectively. Investigate if by changing that value, the efficiency changes in turn

Sir-Bip-Bop commented 1 year ago

It seems like the changes of gap junction strength have no influence on the efficiency of the models, at least that's the conclusion from the comparison between the models after setting all gap junction strength to 0.012

Sir-Bip-Bop commented 1 year ago

[x] Compute how many points does it take to complete one wave fluctuation & take two periods.
[x] look at the voltage plots of the higher frequencies
[x] take only a positive current (each sine wave should have the same response (around 10mV)

Sir-Bip-Bop commented 1 year ago

after changing the three points above, there is still not anything worthy of note that changed, maybe the point computing for the wave fluctuation was wrong

Sir-Bip-Bop commented 1 year ago

Update from March 8th meeting:

[x] Run the simulation for a longer time to be sure to work with a large enough number of oscillations.
[ ] Check if the simulations have an 'ideal' frequency.
[ ] Look at the spike periods in function of intensity and if the slopes are equal
[ ] Why are the models different?

Sir-Bip-Bop commented 1 year ago

Update from March 29:

[x] Make sure the voltage is at steady state
[x] Run simulations for longer (10 periods and take measurements of half)
[x] Increase the frequency to values higher than 100 Hz
[x] Voltage traces of the sine wave

Sir-Bip-Bop commented 1 year ago

Update from April 11/12th:

[x] Re-do the phase delay algorithm
[x] Keep increasing the frequency
[ ] Compare voltage traces of different models at set frequencies
[x] Lower the drive of the HH model
[ ] Try to find out if the problem is the relative amplitude

Sir-Bip-Bop commented 1 year ago

Update from April 20:

[ ] Efficiency at high values of intensity might result in numerical errors for the ML model, check with the original mean. Think of a way to fix.
[x] Try to match the signals completely at high frequencies
[ ] Plot the absolute amplitude ration, instead of the relative one, too see if that changes anything.

Sir-Bip-Bop commented 1 year ago

Update from April 25/26:

[x] The fact that the IZH model is asymmetric could be the cause of the problem. Check what happnes when both of the values are euqal
[x] Study the behvaiour of ML model

Sir-Bip-Bop commented 1 year ago

Update from May 10:

[x] Make sure the presented efficiency is in the steady state