An Agent Oriented Software Engineering Approach for the Adult Stem-Cell Modeling, Simulation and...

An Agent Oriented Software Engineering Approach for the Adult Stem-Cell Modeling, Simulation and Visualization

Seminário de PesquisaMaíra Gatti10/08/2007

© LES/PUC-Rio

Agenda

• Motivation

• The Research Hypothesis and the MAS adequacy

• Related Work and the MAS advantages

• An Agent-based Approach to Model and Simulate Stem-Cell

– The Stem-Cell Self-Organization Description

– The Cell Life Cycle Scope

– The Agent-based Model

– The Agent-based Simulation

• Discussion

– COOPN

• Conclusions and Future Works

• Acknowledgements

© LES/PUC-Rio

Motivation

• How do stem cells behave in the human body?

• How to predict about how and why stem cells behave either individually or collectively?

• Formal models (for instance)

– ordinary differential equations (Lei & Mackey, 2007)

– cellular automata (Agur et al, 2002)

© LES/PUC-Rio

Motivation

• Medicine point of view

– The model serves as a reference, a guide for interpreting experimental results

– The models serves as a powerful means of suggesting new hypotheses

– The simulation lets us test experimentally unfeasible scenarios and can potentially reduce experimental costs

• Software engineering

– How can we model such a large and complex system?

– How to minimize the complexity of the design and implementation of simulation ?

© LES/PUC-Rio

Motivation

• Agent-Oriented Software Engineering (AOSE) and Multi-Agent-Based Simulation (MABS) provides the best of both worlds:

– clean design in the modeling phase

– efficient numerical routines in the simulation phase.

• Current Drawback

– Model’s semantics

– Model and program reuse is limited

• Such dynamic structures can be intuitively represented and efficiently implemented in agent-oriented simulators.

© LES/PUC-Rio

The Research Hypothesis and the MAS adequacy

Aneuploidy => Cancer

LANDIC - UFRJ

© LES/PUC-Rio


• Stem cells are (Loeffler & Roeder, 2002 ):

– a potentially heterogeneous population of functionally undifferentiated cells,

– composed of multi-cellular organisms.

• capable of:

– homing to an appropriate growth heterogeneous environment;

– proliferation;

– production of a large number of differentiated progeny;

– self-renewing or self-maintaining their population;

– regenerating the functional tissue after injury with

– flexibility and reversibility in the use of these options.

© LES/PUC-Rio


• Basic agent’s concepts and capabilities (Jennings, 2000)

– Autonomy

– Interactivity

– Adaptation

– Suitable abstractions

– Flexibility for modeling globally emergent behavior

– Distribution and heterogeneity

– Open systems

– Self-organization

© LES/PUC-Rio


• MAS is an effective way to understand how stem cells organize themselves, and to deal with it emergent global behavior

• The agent-based simulation

– suggests how tiny changes in individual stem cell behavior might lead to disease at the global through the emergent behavior,

– allows temporal analysis,

– reduce costs and risks, and

– could avoid some ethical issues.

© LES/PUC-Rio

Related Work and the MAS advantages

• Existing approaches have been taken and recast in the agent-based modeling and simulation framework, which has demonstrated a number of clear advantages of the agent approach over existing approaches (d’Inverno & Saunders, 2005)

• Mathematical models (Agur et al, 2002, Lei & Mackey, 2007)

– don’t allow expressing partial information about a system, i.e. to formally describe open systems

– explosion of differential equations to model it

– absence of an abstraction for the models

© LES/PUC-Rio

Related Work and the MAS advantages

• Monte Carlo methods (Metropolis & Ulam, 1949)

– probabilistic dependent

– as learning and adaptation are not possible to implement

• Agent-based formal models

– (Theise & d’Inverno, d’Inverno & Prophet, d’Inverno & Saunders, 2003, 2004, 2005)

– the stem cell behavior modeled was

• too simple,

• feasible and

• not adequate or evolvable to an adequate model for the physicians’ experiments

© LES/PUC-Rio

An Agent-based Approach to Model and Simulate Stem-Cell

• The Stem-Cell Self-Organization Description

• The Cell Life Cycle Scope

• The Agent-based Model

• The Agent-based Simulation

© LES/PUC-Rio

The Stem-Cell Self-Organization Description

• The niche

– a specialized cellular environment

– provides stem cells with the support needed for self-renewal

– contains the cells and proteins that constitute the extra cellular environment

• The niche has regulatory mechanisms:

– It saves stem cells from depletion

– It protects the host from over-exuberant stem-cell proliferation

© LES/PUC-Rio

The Stem-Cell Self-Organization Description

© LES/PUC-Rio

The Cell Conceptual Model

• The most actives components during the cell life-cycle

• The mitosis division

– which is the stem cell division during the self renew process

© LES/PUC-Rio

The Cell Conceptual Model

© LES/PUC-Rio

The Cell Life Cycle Scope

© LES/PUC-Rio

The Agent-based Model: MAS-ML Modeling

• Structural Elements– Environment

– Main Organization: Niche

– Organizations: Cell, StemCell, ProgenitorCell, DifferentiatedCell, Cytoplasm

– Objects: CellMembrane, NuclearMembrane, Nuclei, Chromosome, Chromatid, Centrosome, Centriole, Substances, Protein, Organelles, Centromere, Kinetochore, Microtubule, Astral Microtubule, Polar Microtubule, Kinetochore Microtubule

• Static Diagram– Organization Diagram

– Role Diagram

– Class Diagram

• Dynamic Diagram– Sequence Diagram

• Agent interactions

• Goals, plans and actions

© LES/PUC-Rio

The Agent-based Model: MAS-ML Modeling

© LES/PUC-Rio

The Agent-based Simulation

• Developed with Java

• We developed a small framework with – a set of functionalities as send and receive messages,

– the instantiation of the environment, and

– the hierarch structure between agent, role, organization, goal, plan and action

• Visualization tool goals– offer to the users a visual environment by means of which is

possible to follow the Macro and Micro levels of the simulation of stem cell’s cellular life cycle in the niche;

– perceive the difficulties of implementation of the proposed model;

– validate initially the model, verifying if the simulated behaviour in the prototype has similarities with the behavior of the real entities (stem cell’s); and

– compare the semantics of a tool build based on MAS approach to this domain with tools build in other approaches.

© LES/PUC-Rio


• The cells and the images associated to them in each phase

© LES/PUC-Rio


• The cells and the images associated to them in each phase

Macro level

Cell Data Interface

Micro level

Internal Process Interface

© LES/PUC-Rio

Discussion

• COOPN (Concurrent Object-Oriented Petri Nets)

– (Biberstein et al, 2001)

– object-oriented specification language based on synchronized algebraic Petri nets

– defines Petri nets and coordination between Petri nets using object-oriented approach

– Obs: with Petri Nets it was unfeasible

• Comparisons

– Much easier and fast than building the agent-based solution

– However,

• it was not flexible and suitable for the physicians needs

– For instance, adjusting parameters and observe emergent behavior

• Or, not feasible -> explosion of subnets

© LES/PUC-Rio

Discussion

• Our agent-based model solution overloaded the CPU even with few numbers of entities

• the agent-based model solution was not developed considering the visualization process, and the process control

– Hence it was a hard task to incorporate the interface to the framework.

© LES/PUC-Rio

Conclusions

• The stem cell researchers’ collaborators were very excited with the first results

• First emergent phenomenon observed

– the differentiated cells are located at the colony’s extremity while the specialized and stem cells are located at the colony’s centre

© LES/PUC-Rio

Future Works

• A review and a set of tests on the proposed model

• Produce reports about Macro and Micro levels to support the comprehension of the information visualized during the simulation process

© LES/PUC-Rio

Future Works

• Elicit and design the underlined optimized self-organization mechanism

– When does occur a symmetric/ asymmetric division?

• Activators /inhibitors

• Differentiation and self-renew process optimization

• Potentiality influence

• Probabilistic influence

– Self-organization evaluation

• Differentiation and self-renew rate optimum

• Learning (?)

– Observable behavior vs. Simulation

• How to identify local rules in a way that the perturbation induce a cancer?

• How to identify a cancer in the emergent behavior?

– HOWEVER...

© LES/PUC-Rio

Self-Organizing MAS Software Engineering

• How to elicit adaptation’s requirements?

• How to design adaptation mechanisms?

– Learning

• Adaptation patterns

• Adaptation infrastructures

• How to measure adaptation features?

© LES/PUC-Rio

References

• Lei, J., Mackey, M.C. (2007). "Stochastic differential delay equation, moment stability, and appplication to hematopoietic stem cell regulation system", SIAM J. Appl. Math. (2007), 67(2), 387–407.

• Agur, Z., Daniel, Y. and Ginosar, Y. (2002). The universal properties of stem cells as pinpointed by a simple discrete model. Mathematical Biology, 44:79–86, 2002.

• Loeffler, M. and Roeder, I.. Cells Tissues Organs, 171(1):8-26, 2002.• d’Inverno, M. and Saunders, R. (2005). Agent-based modelling of stem cell

organisation in a niche. In Engineering Self-Organising Systems, volume 3464 of LNAI. Springer, 2005.

• Theise, N. D. and d’Inverno, M. (2003). Understanding cell lineages as complex adaptive systems. Blood, Cells, Molecules and Diseases, 32:17–20, 2003.

• d’Inverno, M. and Prophet, J. (2004). Modelling, simulation and visualisation of adult stem cells. In P. Gonzalez, E. Merelli, and A. Omicini, editors, Fourth International Workshop on Network Tools and Applications, NETTAB, pages 105–116, 2004.

• d’Inverno, M. and Saunders, R. (2005). Agent-based modelling of stem cell organisation in a niche. In Engineering Self-Organising Systems, volume 3464 of LNAI. Springer, 2005.

FIM

Maíra Gatti

[email protected]

An Agent Oriented Software Engineering Approach for the Adult Stem-Cell Modeling, Simulation and...

Documents

Transcript of An Agent Oriented Software Engineering Approach for the Adult Stem-Cell Modeling, Simulation and...