Transcript Presentation

Simulation of
Immune System
Answering Questions on the
Natural Immune System Behavior
by Simulations
1
INTRODUCTION
• Three questions :
 Q1: Is the immune innate system able to solve
all attacks?
 Q2: What is the actual role of the adaptive part?
Is it to improve immune reaction that can be
resolved by innate part? Is it to defend against
attacks unresolved by innate part?
 Q3: Quantitative aspect of the intervention of
each part.
2
INTRODUCTION
I. Simulator Choice
CImmSim
II. Simulator Description
II.A The Computational Model
II.B The Entity Description
II.C The interactions
3
INTRODUCTION
III. Simulator Behavior
III.A Basic Step of Algorithms
III.B The interaction Model
III.C The Mutation
IV. The usefulness of CImmSim for Us
V. Illustration of the Simulator Use
CONCLUSION
4
I. SIMULATOR CHOICE
• SIMMUNE :
 only immune adaptive system simulated
 behavior of the cells can be defined with
flexibility
 copy will be available after code update
• SIMISYS :
 not finished project
5
I. SIMULATOR CHOICE
• SIS-I,SIS-II :
 only T and B cells simulated
• IMMSIM:
 adaptive part with humoral and cellular
mediated response
 source code available directly (CImmSim)
 Cellular Automaton (CA) based
6
II. SIMULATOR DESCRIPTION
II.A The Computational Model
• Lymph node is mapped onto a bi-dimensional
hexagonal lattice.
7
II. SIMULATOR DESCRIPTION
II.A The Computational Model
• The hexagonal lattice coordinate system.
(x,y) = (u+v*sin(pi/6),u*cos(pi/6))
8
II. SIMULATOR DESCRIPTION
II.B The Entity Description
9
II. SIMULATOR DESCRIPTION
II.B The Entity Description
10
II. SIMULATOR DESCRIPTION
II.C The Interactions
• External interactions occur among cells which
have the same position on the lattice.
v(d) is affinity potential, d is the Hamming
distance, h is the affinity enhance.
11
II. SIMULATOR DESCRIPTION
II.C The Interactions
• The Hamming distance is the number of
complementary bits between two bit string.
In this case, the Hamming distance is the
length string, i.e d=16.
12
II. SIMULATOR DESCRIPTION
II.C The Interactions
• External interactions occur among cells and
molecules which have the same position on the
lattice.
• Internal interactions inside a cell. (MHC-Ag
peptide interactions).
13
III. SIMULATOR BEHAVIOR
III.A Basic Step of Algorithm
• All the feasible interactions among cells and
molecules take place within a lattice site in a
single time step.
• Diffusion of entities is done at the beginning of
each time step.
• Time step = mitosis cycle = 8 hours.
14
III. SIMULATOR BEHAVIOR
III.B The Interaction Model
15
III. SIMULATOR BEHAVIOR
III.C The Mutation
• In the natural human system, the normal
mutation rate is low.
• In very restricted region of B cells
chromosomes, high rate of mutation occurs:
the somatic hypermutation.
• The somatic hypermutation allows to obtain a
collection of B cells whose receptors are better
to recognize an antigen : affinity maturation
16
III. SIMULATOR BEHAVIOR
III.C The Mutation
• In CImmSim, we can set a parameter which
allow to enhance the affinity to the antigen
only by hypermutation (see later).
17
IV. THE USEFULNESS OF
CIMMSIM FOR US.
• CImmSim does not implement the innate
system.
• CImmSim simulates well the adaptive system.
• We can look at the time taken to clear the
pathogen.
• We can learn, reuse, improve the model in
order to extend the simulator such as be able to
simulate the innate and the adaptive part
together.
18
V. ILLUSTRATION OF THE
SIMULATOR USE.
• The EP cells are killed by Tk cells.
19
V. ILLUSTRATION OF THE
SIMULATOR USE.
• The EP cells are killed by Tk cells.
20
Th and B cells interactions
21
Th and B cells interactions
22
V. ILLUSTRATION OF THE
SIMULATOR USE.
• The minmatch parameter.
23
V. ILLUSTRATION OF THE
SIMULATOR USE.
• The minmatch parameter.
•
!! Drawbacks!!
24
V. ILLUSTRATION OF THE
SIMULATOR USE.
• The Hole parameter.
25
V. ILLUSTRATION OF THE
SIMULATOR USE.
• The Bystander parameter.
26
V. ILLUSTRATION OF THE
SIMULATOR USE.
• The Bystander parameter.
27
V. ILLUSTRATION OF THE
SIMULATOR USE.
• The Bystander parameter.
• Drawback : to much of useless cells
28
CONCLUSION
• The distribution in the time of the project:
 2/7 used for choosing the simulator
 3/7 used to full understand the model and the
simulator behavior
 2/7 used for visualization and the experiments
• Skills exercised:
 theoretical comprehension
 code comprehension
 link between both
29
CONCLUSION
• We can use the well-modeled adaptive system
to see the impact of some parameters.
• We can also retain the model used to simulate
the adaptive part, the interactions system,
thanks to the concept of entities and cellular
automaton.
• Visualization is very important. We have to
know also where are the molecules and to see
the space dependences between entities.
30
CONCLUSION
• Future works would be to model a immune
system simulator able to highlight the manner
of the innate system interacts with the adaptive
system using the feature of CImmSim.
• We can add entities of innate part, create new
interactions between these new entities and the
old one.
• We can create new states for the adaptive cells,
like a “Waiting” state during which the
lymphocytes wait the innate part signal.
31
CONCLUSION
• We have to add new entities as self entities but
not implicated in the immune process in order
to check if the system simulated is dangerous
for self entities.
• Again, visualization is important! We have to
be able to trace all the entities : cells of the
both system and molecules (have a structure to
record the molecule position) and see how they
interact.
32
CONCLUSION
• We have to be able to see the cytokines
secretion between cells which is not possible
only with graph.
• We have to be able to set the influence of each
cell in order to have quantitative aspect of the
intervention of each part.
33