computational science, multidisciplinary research projects; modelling and simulations; morphogenesis;

1. My Computational Science Research
18 July 2014
Nol Chindapol, PhD Candidate
Section Computational Science
University of Amsterdam
Science Park 904, C3156
1098 XH Amsterdam

2. About Me
 Mathematics &
Computer
Engineering
backgrounds
 Passion in
Computational
science

3. Research Projects
 Flow & Morphological Plasticity of Coral
Growth
 Re-Modelling of Purkinje Cells Forest

4. Research questions
 Is morphological plasticity the emergent property of the
external stressors induced by flow constraint to the sessile
organisms?
 What is the interaction between those constraints and
intrinsic property of the organism that leads to self-generation
property of the growth forms?
 How to quantify plastic response relevant to such
constraints?

5. Accretive growth model is coupled with FEM
modelling software (COMSOL) in order to
investigate the plasticity of the resulting growth
forms under uni-directional & bi-directional flow
(Tali et al. 2011)
Data Acquisition

6. Unidirectional Flow & Accretive Growth
Model
Nutrient
Injection
C = 1 mol/m3
Object and
Ground = Sink;
C = 0
Schematic diagram of the simulation (A) A spherical objected represents a simulated object in a first
growth step. (B) A simulation phase involves solving the Navier-Stokes equations and the Advection-
Diffusion equation. (C) Simulated growth form: the accretive growth process generates new growth
layers on top of the previous one.

7. Bidirectional Flow & Symmetry
perseverance Hypotheses
Schematic diagram of the bi-directional flow simulation coupled with the accretive growth model: (a)
initialization phase, (b) the simulation phase consists of two subsequence flow simulation steps. . (c) The
solutions of the nutrient transport are acquired and translocated on the surface of the simulated corals. The
next growth layer is built on top of the previous one by the local growth function

8. Advance Morphometric
Morphometric traits used in our quantitative analysis; (a) local morphometric traits (LMT) are
defined as local traits that are not associated with directional bias e.g. branch spacing (br_spacing),
branch angle (br_angle), ground angle (g_angle), and diameter of branches (da, db and dc) whereas
(b) symmetric-oriented traits (SOT) are those associated with directional bias (h_angle, v_angle, and
spd_angle) i.e. requiring the reference axis.

9. Bifurcations in nature are locally flat –
using data from Neurons & Corals
We do not touch upon this.
Yihawa et al. 2012

10. Morphospace of the flow-induced forms
An overview of the morphospace showing the transition from compact colony to thin branching form
by means of intrinsic model parameter n, while exposed to the flow condition with increasing Pe
number (i.e. decreasing diffusivity D). Red arrows indicate directional variation of flow. . (a) In silico corals
group 1 (b) In silico corals group 2(c) In silico corals group 3

11. Unpublished Work in Turbulent
Flow Simulation
Solved high Reynolds number flow by
using the one equation Spalart-
Allmaras turbulent model, and coupled
with growth model.

12.  Flow & Morphological Plasticity of Coral Growth
 Re-Modelling of Purkinje Cells Forest

13. Research Questions
 How neuron’s complex morphology is created during nervous
development, and eventually leads to the development of
neuron forest.
 How the interaction of genes mediates dendrite self-avoidance by
means of repulsive signal – discriminates self/non self.
 The role of traveling waves in Purkinje cells during early
developmental stage and their significance in cortical microcircuit
wiring.

14. Dendritic infrastructure
 We simulate the
growing micro tubes
i.e. the mechanic
property (e.g.
contact-mediated) is
reduced to a non-
volumetric
abstraction.
 With VTKLines and
Tube Filter

15. Start
Fint
Contact-
mediated
Spatial
Gradient
Branch
Re-
model
Global
Maturation
check
For each neuron
Fext
L
stochastic
End
Update Geometry
 Li = A set of growth event
 Li = Gi (N)
 Gi = FIntrinsic + Fextrinsic
Growth
Mechanisms at
each time step

16. Branching and remodeling
definition
 Let dendritic tree composes of a n number of
cell (n=0,…k), in which each cell constitutes of a
finite set Li (i=0,..m)of non-bifurcated line;
branching mechanism is the addition of cellk+1 to
a non-terminated cell k.
Branch
?
Re-
model?
 ; re-modeling, on the other hand denotes an
action of trimming terminal cell, by which the line
Lm is eliminated, or redefined.

17. N=0
N=7

18. Simulation of
Neuronal
Forest
 Initialization (spherical
objects N=100 at
rand(position)
 Neurite elongation
 Neurite branching (1,2,3
chem) and dendritic shape
formation prob(gradients)
 Stopping criteria (rule based)

19. Thank You!
Please Check coral growth movie 1
Movie 2