Untangling complex syste.., p.97

Untangling Complex Systems, page 97

 

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  

  How to Untangle Complex Systems?

  493

  Philosophy

  and

  Geology

  psychology

  and

  Engineering

  astronomy

  Social

  sciences

  Economy

  Complexity

  Medicine

  Computer

  science

  Biology and

  related

  disciplines

  Mathematics

  Physics

  Chemistry

  FIGURE 13.27 A multidisciplinary approach to face Complexity.

  • Humbleness and wisdom. In fact, science alone cannot furnish a code of morals or a basis

  for aesthetics. Science cannot “furnish the yardstick for measuring, nor the motor for con-

  trolling, man’s love of beauty and truth, his sense of value, or his convictions of faith”

  (Weaver, 1948).

  The polymath figures formed at the interdisciplinary academic degrees in Complexity should

  work along with specialized scientists to build vibrant research teams. The lively multidisciplinary

  research teams will have a much higher potential for facing the Complexity Challenges than the

  potential of teams composed only of members specializing in one particular discipline. In fact, con-

  nections among different disciplines will blossom new great ideas and spark flashes of brilliance.

  However, a fundamental ingredient for success is the willingness of each member to sacrifice selfish

  short-term interests to bring about improvements for all.

  13.5 LAST MOTIVATING SENTENCES PRONOUNCED

  BY “IMPORTANT PEOPLE”

  I want to conclude this breathtaking journey exploring Complexity by citing some motivating sen-

  tences pronounced by eminent philosophers, scientists, and writers.

  He who knows the happiness of understanding

  has gained an infallible friend for life.

  Thinking is to man

  what flying is to birds.

  Don’t follow the example of a chicken

  when you could be a lark.

  Albert Einstein (1879–1955 AD)

  Knowledge is like a sphere.

  The greater it’s volume

  the larger it’s contact with the unknown.

  Blaise Pascal (1623–1662 AD)

  494

  Untangling Complex Systems

  The scientific man does not aim at an immediate result.

  He does not expect that his advanced ideas will be readily taken up.

  His work is like that of the planter—for the future.

  His duty is to lay the foundation for those who are to come, and point the way.

  Nikola Tesla (1856–1943 AD)

  Learn from yesterday,

  live for today,

  hope for tomorrow.

  The important thing is not to stop questioning.

  Albert Einstein (1879–1955 AD)

  One doesn’t discover new lands

  without consenting to lose sight of the shore.

  André Gide (1869–1951 AD)

  13.6 KEY QUESTIONS

  • Which are the principal strategies to tackle the Complexity Challenges?

  • Which are the fundamental elements of an electronic computer and how does it work?

  • What is the future of the electronic computers?

  • What is Natural Computing?

  • Present the three-step methodology to study Complex Systems?

  • Which are the essential attributes of a living being?

  • Explain the basic idea of membrane computing.

  • Explain the Adleman’s idea and procedure of using DNA hybridization reaction to solve

  the Travelling Salesman Problem.

  • Compare DNA with RNA computing.

  • Which are the potentialities and limits of DNA as a memory?

  • Which are the factors that rule biological evolution and are suitable to process information?

  • Describe the essential features of the Human Immune System, which are used to make

  computations.

  • Why are Cellular Automata good models of Complex Systems?

  • Indicate what distinguishes a brain from a Von Neumann computer as far as their compu-

  tational capabilities are concerned.

  • How does Fuzzy logic mimic human ability to compute with words?

  • Which are the routes for developing Artificial Intelligence?

  • How do proteins compute?

  • Describe some methodologies to model phenomena involving Complex Systems.

  • Discuss the potentialities of those algorithms that are based on Swarm Intelligence.

  • Present the Prisoner’s dilemma and its usefulness.

  • Which are the conditions required to compute with a particle in a force field that generates

  a potential energy profile with two wells?

  • Is any irreversible computation an irreversible transformation?

  • Describe the story of mechanical computing.

  • What kind of logic can be processed by using subatomic particles, atoms and molecules?

  13.7 KEY WORDS

  Big Data; Transistor; Von Neumann architecture; Natural Computing; Artificial life; Systems biol-

  ogy; Systems chemistry; Synthetic biology; Artificial Intelligence; Agent; Prisoner’s dilemma;

  Reversible and irreversible logical operations; Quantum logic; Fuzzy logic.

  How to Untangle Complex Systems?

  495

  13.8 HINTS FOR FURTHER READING

  • For those who are interested particularly in Systems Biology, Kitano (2002) suggests visit-

  ing two websites. First, the website of Systems Biology Markup Language (SBML) (http://

  sbml.org/Main_Page) that offers a free and open interchange format for computer models of biological processes. Second, the website of the CellML project (https://www.cellml.

  org/) whose purpose is to store and exchange computer-based mathematical models at a range of resolutions from the subcellular to organism level.

  • For those interested in the theory of Complexity in Medicine, it is worthwhile reading the

  book edited by Sturmberg (2016).

  • A survey on nucleic acid-based devices is the review by Krishnan and Simmel (2011).

  More about DNA origami and their use in chemical robots can be found in the review by

  Hong et al. (2017).

  • More about the thermodynamics of computation in the review by Parrondo et al. (2015).

  • To learn more about Information and have more cues to develop the new theory for under-

  standing Complexity, it is really useful to read the book by von Baeyer (2004).

  • For learning more about the origin of great new ideas, it is nice to read Where Good Ideas

  Come From: The Natural History of Innovation by Johnson (2010).

  • For acquiring an interdisciplinary vision on Complex Systems, it is worthwhile reading

  Humanity in a Creative Universe by the Kauffman (2016). Kauffman sorts through the

  most significant questions and theories in biology, physics, and philosophy.

  13.9 EXERCISES

  13.1. According to the IBM website https://www.ibm.com/analytics/us/en/big-data/, Big Data will amount to 43 Trillion Gigabytes by 2020. If we imagine storing all these data in DNA,

  how much DNA do we need? The data density in bacterial DNA is 1019 bits/cm3 (Extance

  2016). The average density of DNA is 1.71 g/cm3 (Panijpan 1977). Remember that 1 byte

  corresponds to 8 bits.

  13.2. In Xenopus embryos, the cell cycle is driven by a protein circuit centered on the cyclin-

  dependent protein kinase CDK1 and the anaphase-promoting complex APC (Ferrell et al.

  2011). The activation of CDK1 drives the cell into mitosis, whereas the activation of APC

  drives the cell back out of mitosis. A simple two-component model for the cell-cycle oscil-

  lator is shown in Figure 13.28.

  Use a Boolean analysis, to predict the dynamics of the system.

  13.3. A more complex model to describe the cell cycle of Xenopus embryo (see the previous

  exercise) consists of three components (Ferrell et al. 2011). Besides CDK1 and APC, there

  is a third protein that is Polo-like kinase 1 (Plk1), as shown in Figure 13.29.

  Plk1 (P) is activated by CDK1 (C), and, in turn, it activates APC (A). APC inhibits

  CDK1. Use a Boolean analysis, to predict the dynamics of the system.

  CDK1

  APC

  FIGURE 13.28 Mutual relationship between CDK1 and APC.

  496

  Untangling Complex Systems

  APC (A)

  CDK1

  Plk1 (P)

  (C)

  FIGURE 13.29 Model of the cell cycle for the Xenopus embryo.

  13.4. Open NetLogo software (Wilensky 1999). Go to Models Library. Select Biology and

  open the “Ants” Model (Wilensky 1997). Fix “Populations” to 200 and “Diffusion rate”

  to 20.

  For each pheromone “evaporation-rate” value reported below, run the model five times.

  For each run, record the number of ticks it takes until all the food is eaten. Remember

  to click “setup” before each run. Then, average these five numbers. The result is the

  “average time taken” as a function of the evaporation rate value. What is the value of the

  evaporation-rates, reported below, which makes the ants quickest (on average) to eat all

  the food?

  A: “evaporation-rate” = 0

  B: “evaporation-rate” = 5

  C: “evaporation-rate” = 20

  13.5. This exercise is an ideal model of inductive reasoning (Arthur 2015). One hundred people

  decide each week independently whether to go, or not to go, to a bar that offers enter-

  tainment on a certain night. Space is limited. The night is enjoyable if the bar is not too

  crowded, no more than sixty people. There is no sure way to tell the number of people

  coming in advance. Therefore, every agent goes if he expects fewer than sixty to show up or

  stays home if he expects more than sixty to go. There is no prior communication among the

  agents. The only information available is the numbers who went in the past M weeks. This

  model was inspired by the bar El Farol in Santa Fe, which offers Irish music on Thursday

  nights (Wilensky and Rand 2015). Let N be the number of strategies (or mental models)

  each agent has to make predictions. Let t be the current time (i.e., week). The previous

  weeks are thus t − 1, t − 2, and so on. Let A( t − 1) be the attendance at time t − 1, A( t − 2) that at time t − 2, and so on. Each strategy predicts the attendance at time t, S( t), by the following equation:

  S( t) = 100  w 1 A( t − )

  1 + w 2 A( t − 2) +…+ w

  

  M A( t − M) + c

  

  where wi ∈[− ,

  1 + ],

  1

  i

  ∀ = ,

  1 …, M.

  The

  N strategies differ in the values of the coefficients wi. The values of the coefficients

  wi are randomly initialized. The attendance history (previous M time steps) is initialized at

  random, with values between 0 and 99. At each time step t, after making his decision and

  learning the current attendance A( t), each agent evaluates his best current strategy S* as the one that minimizes the error:

  Error( S) = S( t) − A( t) + S( t − )

  1 − A( t − )

  1 +…+ S( t − M) − A( t − M)

  S* will be used by that agent to make his decision on the next round. If S*( t + 1) > 60 (i.e., the overcrowding threshold), the agent does not go; otherwise, he goes.

  How to Untangle Complex Systems?

  497

  (I)

  (II)

  ( i)

  ( f)

  FIGURE 13.30 Mixing of two ideal gases (I) and free expansion of a perfect gas (II) where the state labeled

  as ( i) and ( f) represent the initial and final states, respectively.

  Open NetLogo software (Wilensky 1999). Go to Models Library; select Social Science

  and open the “El Farol” Model (Rand and Wilensky 2007). Find a combination of

  memory-size and number of strategies that reduce the frequency and the extent of over-

  crowding. How does the attendance behave dynamically over time?

  13.6. Determine the variation of the thermodynamic entropy and the information entropy for the

  following two processes: (I) mixing of two ideal gases and (II) free expansion of a perfect

  gas from volume Vi to Vf = V

  2 i (see Figure 13.30).

  13.10 SOLUTIONS OF THE EXERCISES

  13.1. The Big Data will amount to 4 3×1022 bytes = 3 44 ×1023

  .

  .

  bits. If we imagine storing all

  these data in DNA, the volume needed would be

  3.44 1023 bits

  V ( DNA) =

  ×

  = 3.44×104

  3

  cm

  1019 bits/

  3

  cm

  The mass of DNA would be

  .

  3 44 ×104

  3

  cm

  m( DNA) =

  = 2×104g = 20 kg

  g

  .

  1 71

  3

  cm

  It is worthwhile noticing that the Big Data are constantly manipulated to extract useful

  information from them. Therefore, their storage in DNA is not the best choice because

  writing and reading information in DNA is not a fast process, yet. However, DNA requires

  much less volume than a hard disk and flash memory. In fact, the data density of a hard disk

  (based on magnetism) and flash memory (based on electric charges; the name “flash” refers

  to the fact that the data can be erased quickly) are one million and one thousand smaller

  than that of DNA, respectively (Extance 2016).

  13.2. According to the Boolean analysis, the two variables, CDK1 and APC, are either entirely

  ON or entirely OFF. Then, the two-component system has four possible states, indicated in

  Figure 13.31.

  Let us imagine that the system starts from a state wherein it prepares for the process

  of cell division: CDK1 and APC

  (state 1 in Figure). In the first discrete time step,

  OFF

  OFF

  498

  Untangling Complex Systems

  (4)

  (3)

  CDK1: OFF

  CDK1: ON

  APC: ON

  APC: ON

  CDK1: OFF

  CDK1: ON

  APC: OFF

  APC: OFF

  (1)

  (2)

  FIGURE 13.31 Possible binary states of CDK1 and APC.

  CDK1

  APC

  CDK1

  0

  −1

  APC

  +1

  0

  FIGURE 13.32 Connection matrix for the CDK1-APC system.

  CDK1 turns ON because APC is OFF: the system moves from the state 1 to the state 2.

  Then, the active CDK1 activates APC; thus, the system goes from the state 2 to the state 3.

  Now, the active APC inactivates CDK1, and the system goes from the state 3 to the

  state 4. Finally, in the absence of active CDK1, APC becomes inactive, and the system goes

  from the state 4 to the state 1, completing the cycle. The system will oscillate indefinitely.

  The transitions of the system can also be determined by using the “connection matrix”

  reported in Figure 13.32 and representing the state of the two-component system through

  2 × 1 vectors. Here, −1 corresponds to OFF, whereas +1 to ON.

  Ferrell et al. (2011) have demonstrated that if we use ODEs to describe the dynamics

  of this two-component system, we find that it gives damped oscillations to a stable steady

  state.

  13.3. For the three-component system, there are 23 possible states (see Figure 13.33). If we start with the variables in their OFF state (i.e., the state 1), we get periodic cycling of

  six states (from 1 to 6). States 7 and 8 are not visited by the periodic cycle. However, they

  feed into the cycle (see grey arrows) according to the relations existing between the vari-

  ables. Thus, no matter where the system starts, it will converge to the cycle traced by the

  black arrows. The cycle is stable.

  The evolution of the system can also be calculated using the “connection matrix,” shown

  in Figure 13.34, and 3 × 1 vectors to represent the states of the system.

  The analysis of this three-component system by ODEs gives the same result: a stable

  limit cycle (Ferrell et al., 2011).

  How to Untangle Complex Systems?

  499

  (5)

  C: OFF

  C: ON

  P: ON

  P: ON

  A: ON

  A: ON

  (4)

  (3)

  C: ON

  C: OFF

  P: ON

  P: ON

  A: OFF

  A: OFF

  (7)

  (6)

  (8)

  C: OFF

  C: ON

  P: OFF

  P: OFF

  A: ON

  A: ON

  (1)

  (2)

  C: OFF

  C: ON

  P: OFF

  P: OFF

  A: OFF

  A: OFF

  FIGURE 13.33 Possible states for the system consisting of CDK1 (C), Plk1 (P), and APC (A).

  CDK1

  PlK1

  APC

  CDK1

  0

  0

  −1

  PlK1

  +1

  0

  0

  APC

  0

  +1

  0

  FIGURE 13.34 Connection matrix for the CDK1-P-APC system.

  13.4. The results are listed in Table 13.3.

 

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