0
$\begingroup$

I'm studying this classical paper in nonlinear dynamics in biophysics and I want to understand it properly. I'm stuck at this two-variable problem which I've struggled with for too long now.

$$ \dot M = \frac{1}{1+E^m} - a M $$ $$ \dot E = M - b E $$

Where $ \dot z = \frac{dz}{dt} $. It turns out later that $ m \geq 8 $. Note also that, because it's concentrations of molecules, $ M,E \geq 0 $.

This coupled system of equations has a fixed point $ (M_0,E_0) $ where the nullclines $ \dot M = 0 $ and $ \dot E = 0 $ intersect. This occurs when:

$$ M_0 = b E_0 $$ $$ a b E_0 (1+E_0^m) = 1 $$

So far, so good. Now the article says: "We expand near this point by writing $ M = M_0 + X $,   $ E = E_0 + Y $"

$$ \dot X = - m a^2 b^2 E_0^{m+1} Y - a X + O(Y^2) $$ $$ \dot Y = X - b Y $$

I get why they expand near the fixed point and I get that $ O(Y^2) $ means neglecting higher-order terms. And I think I can derive the equation for $ \dot Y $:

$ \dot Y = \dot {(E-E_0)} = M_0 + X - b (E_0 + Y) = b E_0 + X - b E_0 - b Y = X - b Y $

But I need your help to understand how to get $ \dot X $ just from taylor expansion and the implicit equation for $ E_0 $.

The paper can be found here: http://www.math.us.edu.pl/mtyran/dydaktyka/biomatematyka/griffith_1968_I.pdf J. S. Griffith "Mathematics of Cellular Control Processes" J. Theoret. Biol. (1968)

Thanks in advance.

$\endgroup$
1
$\begingroup$

$$ \dot{X} = \dot{M} = \frac{1}{1+(E_0 + Y)^m } -a(M_0 +X) = $$ $$ \frac{1}{1+E_0^m + mE_0^{m-1} Y + O(Y^2) } -aM_0 -aX = $$ $$ \frac{1}{1+E_0^m} \left( 1-\frac{mE_0^{m-1} }{1+E_0^m } Y + O(Y^2) \right) -aM_0 -aX $$ Now use $1/(1+E_0^m) =abE_0 $ as well as $aM_0 = abE_0 $, yielding $$ \dot{X} =-a^2 b^2 m E_0^{m+1} Y +O(Y^2 ) -aX $$ as desired.

$\endgroup$
2
  • $\begingroup$ Thanks! It looks correct. Can I just ask how you split the fraction and got the Y-terms from the denominator to the numerator? $\endgroup$
    – Norregaard
    Jul 23 '20 at 8:00
  • $\begingroup$ From the second to the third line, I factored out $1/(1+E_0^m )$ and then expanded in a geometric series. $\endgroup$ Jul 23 '20 at 13:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.