Untangling complex syste.., p.61

Untangling Complex Systems, page 61

 

Untangling Complex Systems
Select Voice:
Brian (uk)
Emma (uk)  
Amy (uk)
Eric (us)
Ivy (us)
Joey (us)
Salli (us)  
Justin (us)
Jennifer (us)  
Kimberly (us)  
Kendra (us)
Russell (au)
Nicole (au)



Larger Font   Reset Font Size   Smaller Font  

  

  t

  ∂ 2 

   r 

  ∂ y

  

  r

   2 y 

  1

  

  = β

  α

  y 1+

  xy

  xy +

  ∂

  δ

  ∂

  

   +

   2 

  t

  

  β

  

  ∂

   r 

  All the four conditions for having a Turing pattern, considering a diffusion-driven instabil-

  ity from the stationary state ( x

  ) 0 0)

  s , ys = ( ,

  , are verified. In fact,

  I.

  α + β = 0 899

  .

  − 0 9

  . 1 < 0;

  II.

  αβ − γ = ( .

  0

  )(

  899 − .

  0 9 )

  1 − (− .

  0

  )

  899 > 0;

  III.

  δ Dβ +αδ = ( .

  0

  )(

  0021 .

  0

  )

  516 (− .

  0 9 )

  1 + ( .

  0

  )

  899 ( .

  0

  )

  0021 > 0;

  IV.

  δ Dβ +αδ ≅ × −4 >

  δ

  D 2 (αβ −γ ) ≅

  × −

  9 10

  2

  8 6 10 4

  .

  The critical wavenumber is K = [

  D +

  ]1 2 ≅

  −

  (

  )((

  ) (

  )) /

  1 2

  14

  1

  α δ

  β δ

  cm . The wavelength is

  λ = ( π

  2 K ) ≅ 0 449

  .

  cm. Since the spatial domain is L = [−1, +1], its total length is 2. It

  follows that the number of waves in the domain is (2 / λ) ≈ .

  4 5. This can be confirmed

  by solving numerically our system of partial differential equations, using the following

  MATLAB file (adapted from Schneider 2012):

  %Grid size

  Tf=3000;

  a=-1;

  % Lower boundary

  b=1;

  % Upper boundary

  M=100;

  % M is the number of spaces between points a and b.

  dr=(b-a)/M;

  % dr is delta r

  r=linspace(a,b,M+1); % M+1 equally spaced r vectors including a and b.

  %Time stepping

  dt=0.04;

  % dt is delta t the time step

  N=Tf/dt;

  % N is the number of time steps

  300

  Untangling Complex Systems

  %Constant Values

  D=0.516;

  % D is the Diffusion coefficient Du/Dv

  delta=0.0021;

  % sizes the domain for particular wavelengths

  alpha=0.899;

  % a is alpha, a coefficient in f and g

  beta=-0.91;

  % b is beta, another coefficient in f and g

  r1=3.5;

  % r1 is the cubic term

  r2=0;

  % r2 is the quadratic term

  gamma=-alpha;

  % g is for gamma

  %pre-allocation

  xnp1=zeros(M+3,1);

  ynp1=zeros(M+3,1);

  %Initial Conditions

  xn=-0.5+rand(M+3,1); %Begin with a random point between [-0.5,0.5]

  yn=-0.5+rand(M+3,1);

  for n=1:N

  xn(1)=xn(3);

  %Boundary conditions on left flux is zero

  xn(M+3)=xn(M+1);

  %Boundary conditions on right

  yn(1)=yn(3);

  yn(M+3)=yn(M+1);

  for j=2:M+2

  %Source function for x and y

  srcx=alpha*xn(j)*(1-r1*yn(j)^2)+yn(j)*(1-r2*xn(j));

  srcy=beta*yn(j)*(1+(alpha*r1/beta)*xn(j)*yn(j))+xn(j)*(gamma+r2*yn(j));

  Lapx=(xn(j-1)-2*xn(j)+xn(j+1))/dr^2; %Laplacian x

  Lapy=(yn(j-1)-2*yn(j)+yn(j+1))/dr^2; %Laplacian y

  xnp1(j)=xn(j)+dt*(D*delta*Lapx+srcx);

  ynp1(j)=yn(j)+dt*(delta*Lapy+srcy);

  end

  xn=xnp1;

  yn=ynp1;

  % Graphing

  if mod(n,25000)==0

  %subplot(2,1,2)

  plot(r, xn(2:M+2),r, yn(2:M+2))

  axis([-1,1,-1,1]);

  fprintf(‘Time t = %fn’,n*dt);

  input(‘Hit enter to continue:’)

  end

  end

  In such MATLAB file, we apply the Euler’s method (see Appendix A) to a semi-discretized

  Reaction-Diffusion system. Moreover, the second derivative is expressed as finite differences:

  d 2 x

  x( r

  r

  ∆ ) 2 x( r) x( r

  r

  =

  ∆ )

  Lapx =

  −

  −

  +

  +

  = ( xn( j −1) − 2* xn( j) + xn( j + )

  1 ) / dr 2

  dr 2

  ( r 2

  ∆ )

  The initial conditions are fixed as a randomly generated set of values between [−0.5,

  +0.5].

  The result of the calculation is shown in Figure 9.33.

  We notice that the x and y species are spatially distributed in anti-phase conditions:

  in the point where x is at its maximum, y is at its minimum and vice versa. This result

  verifies that in the places where the activator prevails, the inhibitor succumbs and

  vice versa. Moreover, we confirm that the number of waves in the spatial domain is

  about 4.5.

  The Emergence of Order in Space

  301

  0.2

  xy

  sie 0.1

  ec

  0.0

  −0.1

  Abundance of sp

  −0.2−1.0

  −0.5

  0.0

  0.5

  1.0

  r

  FIGURE 9.33 Turing pattern involving x and y in one-dimensional space of coordinate r.

  9.7. If you have MATLAB at your disposal, an example of .m file that can be used to find the

  solution of the exercise is the following one (adapted from Schneider 2012):

  %Grid size

  Tf=100000;

  a=-1;

  % Lower boundary

  b=1;

  % Upper boundary

  M=100;

  % M is the number of spaces between points a

  and b.

  dx=0.04;

  %(b-a)/M; % dx is delta x

  dy=0.04;

  %(b-a)/M;

  x=linspace(a, b,M+1);

  % M+1 equally spaced x vectors including a and b.

  y=linspace(a, b,M+1);

  %Time stepping

  dt=0.08;

  %100*(dx^2)/2; % dt is delta t the time step

  N=Tf/dt;

  % N is the number of time steps in the interval

  [0,1]

  %Constant Values

  D=0.516;

  % D is the Diffusion coefficient Du/Dv

  delta=0.0021;

  % sizes the domain for particular wavelengths

  alpha=0.899;

  % a is alpha, a coefficient in f and g (-a is

  gamma)

  beta=-0.91;

  % b is beta, another coefficient in f and g

  r1=3.5;

  % r1 is the cubic term

  r2=0;

  % r2 is the quadratic term

  gamma=-alpha;

  % g is for gamma

  %pre-allocation

  unp1=zeros(M+3,M+3);

  vnp1=zeros(M+3,M+3);

  %Initial Conditions

  un=-0.5+rand(M+3,M+3); %Begin with a random point between [-0.5,0.5]

  vn=-0.5+rand(M+3,M+3);

  for n=1:N

  for i=2:M+2

  un(i,1)=un(i,3);

  %Boundary conditions on left flux is zero

  un(i,M+3)=un(i,M+1);

  %Boundary conditions on right

  vn(i,1)=vn(i,3);

  vn(i,M+3)=vn(i,M+1);

  end

  302

  Untangling Complex Systems

  for j=2:M+2

  un(1,j)=un(3,j);

  %Boundary conditions on left

  un(M+3,j)=un(M+1,j);

  %Boundary conditions on right

  vn(1,j)=vn(3,j);

  vn(M+3,j)=vn(M+1,j);

  end

  for i=2:M+2

  for j=2:M+2

  %Source function for u and v

  srcu=alpha*un(i, j)*(1-r1*vn(i, j)^2)+vn(i, j)*(1-r2*un(i, j));

  srcv=beta*vn(i, j)*(1+(alpha*r1/beta)*un(i, j)*vn(i, j))

  +un(i, j)*(gamma+r2*vn(i, j));

  uxx=(un(i-1,j)-2*un(i, j)+un(i+1,j))/dx^2; %Laplacian u

  vxx=(vn(i-1,j)-2*vn(i, j)+vn(i+1,j))/dx^2; %Laplacian v

  uyy=(un(i, j-1)-2*un(i, j)+un(i, j+1))/dy^2; %Laplacian u

  vyy=(vn(i, j-1)-2*vn(i, j)+vn(i, j+1))/dy^2; %Laplacian v

  Lapu=uxx+uyy;

  Lapv=vxx+vyy;

  unp1(i, j)=un(i, j)+dt*(D*delta*Lapu+srcu);

  vnp1(i, j)=vn(i, j)+dt*(delta*Lapv+srcv);

  end

  end

  un=unp1;

  vn=vnp1;

  % Graphing

  if mod(n,6250)==0

  %subplot(2,1,2)

  hdl = surf(x,y,un(2:M+2,2:M+2));

  set(hdl,’edgecolor’,’none’);

  axis([-1, 1,-1,1]);

  %caxis([-10,15]);

  view(2);

  colorbar;

  fprintf(‘Time t = %fn’,n*dt);

  ch = input(‘Hit enter to continue:’,’s’);

  if (strcmp(ch,’k’) == 1)

  keyboard;

  end

  end

  end

  The patterns achieved when r

  3.5,

  0, and there are not quadratic terms in the PDEs,

  1 =

  r 2 =

  are shown in Figure 9.34. They are made of stripes.

  The patterns obtained when r

  0.02,

  0.2, and the contribution of the cubic terms

  1 =

  r 2 =

  in the PDEs is pretty small, are shown in Figure 9.35. They are made of spots.

  The patterns obtained when r

  3.5,

  0.2 and both the cubic and the quadratic terms

  1 =

  r 2 =

  are relevant, are shown in Figure 9.36. They contain both stripes and spots.

  In synthesis, when there is only the contribution of the cubic terms, we observe stripes; when

  the quadratic term is dominant, we have spots. Finally, when both the quadratic and the cubic

  terms contribute appreciably, we observe patterns with spots and stripes (Barrio et al. 1999).

  The Emergence of Order in Space

  303

  u

  v

  −

  −

  0.1750

  0.2250

  −

  0.8

  0.1344

  −0.1666

  0.8

  −0.09375

  −0.1063

  −0.05313

  −0.04688

  −0.01250

  0.01250

  0.4

  0.02813

  0.07187

  0.4

  0.06875

  0.1312

  0.1094

  0.1905

  y 0.0

  0.2500

  0.1500

  y 0.0

  −0.4

  −0.4

  −0.8

  −0.8

  −0.8

  −0.4

  0.0

  0.4

  0.8

  −0.8

  −0.4

  0.0

  0.4

  0.8

  x

  x

  FIGURE 9.34 Profiles of u (on the left) and v (on the right) obtained when r

  3.5,

  0, and within the

  1 =

  r 2 =

  square box having x and y as spatial coordinates.

  u

  v

  −5.960

  −11.23

  −

  −

  3.344

  9.547

  0.8

  −0.7375

  0.8

  −7.899

  1.869

  −6.191

  4.475

  −4.513

  0.4

  7.081

  −2.834

  0.4

  9.688

  −1.156

  12.29

  0.5219

  y

  14.90

  2.200

  0.0

  y 0.0

  −0.4

  −0.4

  −0.8

  −0.8

  −0.8

  −0.4

  0.0

  0.4

  0.8

  −0.8

  −0.4

  0.0

  0.4

  0.8

  x

  x

  FIGURE 9.35 Profiles of u (on the left) and v (on the right) obtained when r

  0.02,

  0.2, and within the

  1 =

  r 2 =

  square box having x and y as spatial coordinates.

  u

  v

  −0.1740

  −0.1660

  −

  −

  0.1218

  0.1308

  0.8

  −0.09950

  0.8

  −0.09550

  −0.01725

  −0.06025

  0.03500

  −0.02500

  0.4

  0.08725

  0.4

  0.01025

  0.1395

  0.04550

  0.1918

  0.08075

  y

  0.2440

  0.1160

  0.0

  y 0.0

  −0.4

  −0.4

  −0.8

  −0.8

  −0.8

  −0.4

  0.0

  0.4

  0.8

  −0.8

  −0.4

  0.0

  0.4

  0.8

  x

  x

  FIGURE 9.36 Profiles of u (on the left) and v (on the right) obtained when r = 3.5, = 0.2, and within the 1

  r 2

  square box having x and y as spatial coordinates.

  304

  Untangling Complex Systems

  TABLE 9.4

  Six Combinations of the Parameters of the Turing Model and Pictures of the Patterns They

  Generate. In the Pictures, the Activator is Gray, Whereas the Inhibitor is Black

  Name

  au

  bu

  cu

  du

  av

  bv

  cv

  dv

  Du

  Dv

  Pattern

  Spot

  0.08

  −0.08

  0.005

  0.03

  0.1

  0

  −0.15

  0.06

  0.02

  0.5

  Net

  0.08

  −0.08

  0.2

  0.03

  0.1

  0

  −0.15

  0.06

  0.02

  0.5

  Stripe2

  0.08

  −0.08

  0.04

  0.03

  0.1

  0

  −0.15

  0.08

  0.02

  0.5

  Stripe3

  0.08

  −0.08

  0

  0.03

  0.1

  0

  −0.15

  0.08

  0.02

  0.5

  Dapple

  0.15

  −0.08

  0.025

  0.03

  0.15

  0

  −0.15

  0.06

  0.02

  0.5

  Spot2

  0.1

  −0.08

  0.025

  0.03

  0.15

  0

  −0.1

  0.06

  0.02

  0.5

  9.8. The combinations of the parameters available in the RD simulator by Kondo and his team

  are listed in Table 9.4. In the last column of Table 9.4, there are the pictures of the patterns generated by numerical solution of the partial differential equations.

  A new combination of parameters values is the following one:

  au = 0 0

  . 8, bu = −0 0

  . 8, cu = 0 2

  . , du = 0 0

  . 3, Du = 0 0

  . 2, av = 0 1

  . , bv = 0, cv = −0 1

  . 5,

  dv = 0 0

  . 7, Dv = 0 5

  . . This combination differs from that named as “Net” in Table 9.4 only in

  dv. The new combination verifies the four conditions required to have a Turing pattern. In fact,

  The Emergence of Order in Space

  305

  I.

  tr( J ) = au − du + bv − dv = .

  0 08 − .

  0 03 + 0 − .

  0 07 < 0

  II.

  det( J ) = ( a

  )(

  )

  u − du

  bv − dv − buav = ( .

  0 0 )(

  5 −0. )

  07 − (−0.0 )(

  8

  .

  0 )

  1 = .

  0 00445 > 0

  III.

  D (

  )

  (

  )

  u bv − dv + Dv au − du = (0.0 )(

  2 −0. )

  07 + (0. )(

  5

  .

  0 0 )

  5 = .

  0 0236 > 0

  IV.

  D (

  )

  (

  ) 0 0236

  .

  2

  det( )]1/2

  /

  1 2

  [

  ]

  u bv − dv + Dv au − du =

  > [ DuDv

  J

  =2 (0 0

  . 2)(0 5

  . )(0.

  )

  0045

  = .

  0 013

  If we start from a randomized condition, the final stationary pattern is shown in

  Figure 9.37 as graph a. If we start from configuration a and we reduce the values of the

  diffusion coefficient to Du = 0 002

  .

  and Dv = 0 0

  . 5, we obtain the pattern shown in b of

  Figure 9.37. On the other hand, if we use the values Du = 0 002

  .

  and Dv = 0 0

  . 5 from an

  initial randomized distribution of u and v into space, we obtain the pattern labeled as c in

  Figure 9.37.

  If we take the pattern obtained by the combination named as “Stripe2” in Table 9.4 as

 

Add Fast Bookmark
Load Fast Bookmark
Turn Navi On
Turn Navi On
Turn Navi On
Scroll Up
Turn Navi On
Scroll
Turn Navi On
183