The Hum Of The Earth, Science 
A few days after the Loma Prieta earthquake I was in
Santa Cruz horseback riding with my non-scientist friend when she
observed: "This is earthquake weather". I suppressed an arrogant
comment about how seismic and atmospheric processes are not
coupled and I am glad I did because now it seems that they are
indeed connected.
Nishida et al. describe evidence that the global "hum" of the
earth, background free oscillations with oscillations of
amplitudes of 0.5 nano-galileos (that is accelerations that are an
American Trillion (= 1 Million Millions) smaller than what we
experience as terrestrial gravitation) and periods of around four
minutes. That means we cannot really hear or feel those
oscillations that need super precise instruments and a quiet earth
(no earthquakes anywhere) to be detected.
The fact that the earth is still ringing after a couple of
billion years after its creation implies that there must be some
force that keeps it continuously oscillating like a string of a
violin. The hypothesis that earthquakes are the driving force is
not consistent with the purity and regularity of the
earth-hum:
It changes its intensity with a yearly period and is strongest
when it is summer in the Northern hemisphere. This is a strong
hint pointing at atmospheric origins: Since most landmasses are in
the Northern hemisphere, seasonal temperature variations are also
stronger there. The frequency range of this oscillation is also in
resonance with well-known atmospheric oscillations.
The story that Nishida et al. propose is that heating of the
atmosphere leads to pressure changes (thermals) that are
selectively amplified in the resonant atmospheric modes (like in a
laser) which then couple to those seismic modes of the solid earth
that resonate at the same frequency.
It is not very likely, however, that those oscillations will be
strong enough to trigger earthquakes.
-
"Old dogs don't learn new tricks!" The same apparently
also applies to mice and humans and Joe Tsien explains that there
is a biochemical basis for this folklore. Our memories are stored
-according to current brain theories-in the form of different
values of the "synaptic strength" between connected neurons in the
sense of how likely it is that one firing neuron will trigger the
firing of the second neuron. Every time this synchronous firing
takes place, the strength of the synapse between the two neurons
will be either increased ("long term potentiation" (LTP)) or
decreased ("long term depression" (LTD)). This process is known as
"Hebbian learning" (see also ComDig 2000.10.1):
The mechanism for this modification is understood to consist of
a modification in the numbers and properties of NMDA receptors
that sit on the cell membranes of postsynaptic neurons. They are
basically small pores that control the entry of calcium ions into
the neurons. Hebbian learning is implemented in the form that
their modification requires the simultaneous presence of certain
neuro-transmitters (i.e. the pre-synaptic cell must have just
fired) as well as a depolarization of the post-synaptic membrane
(i.e. the second cell must have just fired). This shows that they
work as molecular switches and coincidence detectors as required
by Hebb's learning rule.
How does age come into this game? It turns out that there are
two types of NMDA receptors that have either NR2A or NR2B
subunits. They differ in that the NR2B kind stays open a little
longer, thereby making learning a little easier. It is no surprise
that as we age we produce more of the NR2A kind. Tsien and his
collaborators were able to genetically modify mice so that they
produce more NR2B type receptors when they are adult. And sure
enough, they could be shown to learn faster and remember
longer.
Tsien hopes that within ten years or so this discovery will
lead to drugs that will help patients with modest Alzheimer's
disease. We can also expect a new generation of "brain stimulants"
for people who need to cram for an exam or who look for a short
cut to higher intelligence. The implications for society in any
case have a considerable potential.
-
One of the characteristic properties of complex dynamical
systems is that they appear to violate the second law of
thermodynamics according to which a system left to itself will
change in a way that its entropy or state of disorder increases.
Chemical reactions therefore were expected to simply proceed from
initial components to a final product. Therefore it took many
years before the scientific community accepted chemical
oscillations like the Belousov-Zhabotinski reaction. These
reactions also can form waves and spatial patterns instead of just
diffusing and forming a homogenous mixture.
The solution of the mystery of the apparent violation of
fundamental physical laws is that these reactions share one
characteristic property of living beings, namely they have a
simple metabolism: They can only continue to oscillate or form new
patterns as long as they are "fed" with new chemicals. These
chemical clocks were considered to provide a mechanism for many
biological clocks.
Petty et al. have demonstrated that such chemical oscillations
and waves do indeed form already in individual cells. They studied
the concentrations of chemical compounds (NAD(P)H and also the
acid levels in blood cells and they could observe concentration
waves of these chemicals move across the cell body.
On molecular time-scales these oscillations are very slow,
about three per second, about the rate at which you clap your
hands in applause.
The energy for this oscillation comes from the metabolism of
sugar -glycolysis- a reaction where in vitro oscillations have
been observed in the past.
It would be interesting to see if this newly discovered "cell
clock" is in resonance with other biological clocks.
- A
Finger On The Pulse,
Philip Ball, Nature Science Update,
3/24/00
- Imaging Sustained Dissipative
Patterns in the Metabolism of Individual Living
Cells. Petty, H.R., Worth, R.G. & Kindzelskii,
A.L. Physical Review Letters 84, 2754-2757
(2000)
-
The coordinated movements of athletes are one of the most
profitable examples of outcomes of human complex adaptive systems.
Often the movement involves placing balls of different sizes (e.g.
basket/foot/base/tennis/golf balls) at different locations. The
size of the reward for specific movements is often inversely
correlated with the size of the ball and the total amount of
movement produced.
The challenge is to reduce the number of degrees of freedom in
the system of voluntarily controlled muscles in a way that the
desired movement is accurately produced with high reliability.
Loeb et al. studied how these muscle synergies are produced in
the simpler organism of a frog. They measured the forces produced
by the 65,536 combinations of the 16 limb muscles of the frog's
hind leg. They found that a very small number of those
combinations (as few as 23) are robust with respect to activation
noise: these combinations would stabilize a limb at a predictable
and restricted location even if the forces produced by individual
muscles out of the combination varied significantly. This result
is consistent with the observation that skilled athletes often
show a large variation of movement details but with a consistent
overall outcome.
The researchers measured force fields of the frog ankle and
determined their Convergent Equilibrium points (CEP or "attractive
fixed points" in dynamical systems terminology) under stimulation
of different muscle combinations. They developed a method that
allows them to make predictions about the limb stabilization
properties from the measurement of the force fields of individual
muscles. They also tried to estimate the muscle force from changes
in the EMG a method that could eventually lead to powerful
augmented feedback that could help to improve learning of complex
motor tasks.
-
Aided by a microscope and a digital camera, a team of
researchers led by an Ohio University cell biologist has snapped
the first pictures of a sight that has eluded scientists for 15
years - tiny threads of protein key to the health of the nervous
system darting along nerve fibers.
What they've documented with time-lapse photography could one
day lead to a better understanding of nerve malfunction in Lou
Gehrig's disease and other, similar neurological disorders.
For the past two decades, scientists have struggled to observe
how proteins critical to the growth and maintenance of the nervous
system travel through the body's network of nerves. In the March
issue of the journal Nature Cell Biology, Anthony Brown, Ohio
University associate professor of cell biology, and his colleagues
report on a new technique that allowed them to watch and
photograph the movement of microscopic threads of protein called
neurofilaments in nerve fibers.
A logjam of this neurofilament movement, which blocks other
biological processes vital to the nerve's survival, has been seen
in patients with certain neurological disorders, such as Lou
Gehrig's disease. (…)
The researchers' observation of the neurofilament movement
(...) has provided a rare glimpse of slow axonal transport, the
process by which many of the proteins in the nerve cell's
cytoplasm travel from the nerve cell body along the nerve fibers,
also called axons. These proteins are crucial for the development
and maintenance of axons, branch-like fibers that communicate
information from the nervous system to other areas of the body.
The study suggests that neurofilaments move in fast but
infrequent spurts - at rates of up to two-thousandths of a
millimeter per second. This finding argues against a previous
theory of slow axonal transport, which hypothesized that
neurofilaments and other transported proteins travel in a slow,
steady manner. The long pauses between the quick movements the
team observed may be one reason why scientists have had a hard
time tracking the process, Brown says. (…)
To make the movement visible, Brown's team fused DNA coding for
neurofilament protein with the DNA coding for the protein that
makes jellyfish glow green. But as most nerve fibers are packed
with neurofilaments along their entire length, at first all the
researchers could see was one long bright green strip. The
scientists solved the problem by studying nerve cells that had
fewer neurofilaments, which showed visible gaps in the green
fluorescence. They digitally photographed the movement of the
neurofilaments by waiting for them to sprint across these gaps.
"The gaps are basically like little windows on the cytoplasm of
the axon," Brown says. "They allow us to see movement that we
normally wouldn't be able to see. That really was the key."
Now that the researchers have observed neurofilaments in
transit, Brown's laboratory will begin the study of how the
proteins move. "Only once we understand the mechanism of movement,
can we really start to understand the mechanism that might impair
movement," Brown says.
- Scientists
Capture New Images Of Movement In
Nerves, Science Daily,
Ohio University, 3/22/2000
- See also: Rapid
movement of axonal neurofilaments interrupted by
prolonged pauses,
Lei Wang, Chung-liang Ho, Dongming Sun, Ronald
K.H. Liem and Anthony Brown, Nature Cell Biology, pp
137 - 141
- To view a movie of the nerve
activity, go to http://www.nature.com/ncb//suppl/ncb0300/ncb0300_137/
Yeast, Flies, Worms, and Us, Science
Genomics is the new Big Science of this century. Just as
it was the case in the high energy physics of the seventies and
eighties publications have author-lists that comprise a
significant fraction of the text and results can be predicted like
large-scale industrial developments. We can estimate when the
genome of which organism will be sequenced and what type of
medical and other insights we can expect.
At the same time the cumulative results completely change our
worldview in terms of our place among living beings. As the genome
of the third important eukaryotic model organism has just been
completely decoded, Rubin and more than fifty co-authors compare
their properties among each other and with humans. The organisms
that play a central role in genetic medicine because of their
genetic similarity with humans are a yeast (Saccharomyces
cerevisiae"), a worm (Caenorhabditis elegans), and a fly
(Drosophila melanogaster). The importance of these genetic model
organisms can be illustrated by the fact that out of 289 known
human disease genes the fly has 177 orthologs. That means that
they show significant similarity over at least 80% of their
lengths and can be expected to have similar responses e.g. to
pharmaceuticals. Other differences have a known reason, for
instance flies don't use blood for oxygen transport and
correspondingly don't have any hemoglobin related genes.
Similarities between the different model organisms include the
size of the protein coding parts of the genes. They are similar
for worm and fly and about half as big for yeast. That is quite
surprising since complicated organisms like worms and flies look
much more complex than single-celled yeast. (Note that these sizes
do not take into account the "junk dna", see also: ComDig
2000.6.3 " Why Onions Have More DNA Than You Do").
The researchers are optimistic about future applications:
"For the first time, we can envision obtaining the data needed
to understand the behavior of a complex regulatory network. Of
course, collecting these data is a massive task, and developing
methods to analyze the data is even more daunting. But it is no
longer ludicrous to try. (…)The power and speed of this in
vivo system are unparalleled, and we anticipate the increased use
of such "humanized" fly models. (…) The relative simplicity
and manipulability of the fly genome means that we can address
some of these biological questions much more readily than in
vertebrates."
- Comparative
Genomics of the
Eukaryotes, Gerald M.
Rubin, Mark D. Yandell, Jennifer R. Wortman, George L.
Gabor Miklos, Catherine R. Nelson, Iswar K. Hariharan,
Mark E. Fortini, Peter W. Li, Rolf Apweiler, Wolfgang
Fleischmann, J. Michael Cherry, Steven Henikoff,
Marian P. Skupski, Sima Misra, Michael Ashburner, Ewan
Birney, Mark S. Boguski, Thomas Brody, Peter
Brokstein, Susan E. Celniker, Stephen A. Chervitz,
David Coates, Anibal Cravchik, Andrei Gabrielian,
Richard F. Galle, William M. Gelbart, Reed A. George,
Lawrence S. B. Goldstein, Fangcheng Gong, Ping Guan,
Nomi L. Harris, Bruce A. Hay, Roger A. Hoskins, Jiayin
Li, Zhenya Li, Richard O. Hynes, S. J. M. Jones, Peter
M. Kuehl, Bruno Lemaitre, J. Troy Littleton, Deborah
K. Morrison, Chris Mungall, Patrick H. O'Farrell,
Oxana K. Pickeral, Chris Shue, Leslie B. Vosshall,
Jiong Zhang, Qi Zhao, Xiangqun H. Zheng, Fei Zhong,
Wenyan Zhong, Richard Gibbs, J. Craig Venter, Mark D.
Adams, Suzanna Lewis, Science, Volume 287, Number 5461
Issue of 24 Mar 2000, pp. 2204 - 2215
A Drosophila Model Of Parkinson's Disease, Nature
Flies (to be more precise fruit flies Drosophilae
melanogaster) are the talk of the town especially since their
genome has been completely sequenced recently. Looking at a
genetic level flies look more and more like members of the family
and memories about Franz Kafka's "Metamorphosis" keep coming up.
After all we know that they get tired and go to sleep (see
ComDig
2000.10.5), they have taste and now it seems that in their old
age of fifty days or so they have to worry about Parkinson's
disease just like humans at around the same age (in years). The
symptoms are surprisingly similar for flies and for humans. They
include loss of dopaminergic neurons while the aging brain doesn't
show any signs of widespread degeneration.
On a behavioral level flies show motor dysfunctions just like
humans: for 55 days Feany et al. gently (!) tapped forty fruit
flies to the bottom of a plastic vial. Then they counted how many
of them made it backs to the top of the vial within 18seconds.
Whereas the control flies had no problem to make it out of the
vial up to an age of over thirty, those that were genetically
modified with the human Parkinson protein had already significant
difficulties in their mid-twenties.
Right Brain Essential For Japanese Mirror Reading, Brain
Summary: Functional magnetic resonance imaging
(fMRI) was used to investigate the neural substrates responsible
for Japanese kana mirror reading. Japanese kana words, arranged
vertically in the up-to-down direction, were used in the mirror
reading task in 10 normal right-handed Japanese adults. Since
both mirror reversed and normally oriented kana items are read in
the same (up-to-down) direction, our study could minimize the
oculomotor effects which often occurred in the process of the
mirror reading of alphabetical language.
By using the random effect analysis method of SPM96,
significant increase of the blood oxygen level dependent signal
during mirror reading relative to normal reading was detected in
multiple brain regions, including the bilateral superior occipital
gyri, bilateral middle occipital gyri corresponding to Brodmann's
area (BA) 18/19, bilateral lingual gyri (BA 19), left inferior
occipital gyrus (BA 18), left inferior temporal cortex (BA 37),
bilateral fusiform gyri (BA 19), right superior parietal cortex
(SPC, BA 7), left inferior frontal gyrus (BA 44/45) and an
inferior part of the left BA 6. In addition to these cortical
regions, the right caudate nucleus and right cerebellum were also
activated.
The activation found in the right SPC and the left inferior
temporal region is consistent with the hypothesis that mirror
reading involves both the dorsal visuospatial and ventral object
recognition pathways. In particular, a significant correlation
was found between the fMRI signal change in the right SPC and the
behavioural performance (error index) in the task. This may
reflect increased demand of the right SPC for the spatial
transformation which is required for the accurate recognition of
mirror reversed kana items.
This relationship between the hemodynamic response in a
specific brain area and the behavioural data provides a new
evidence for the essential role of the right SPC in Japanese kana
mirror reading.
- Essential Role Of The Right
Superior Parietal Cortex In Japanese Kana Mirror
Reading, An fMRI Study,Yun Dong, Hidenao Fukuyama,
Manabu Honda, Tomohisa Okada, Takashi Hanakawa,
Kimihiro Nakamura, Yasuhiro Nagahama, Takashi
Nagamine, Junji Konishi, Hiroshi Shibasaki, Brain,
Vol. 123, No. 4, 790-799, April 2000
-
Over the last 15 years, laboratory research has revealed
a high level of intelligence in the bottlenosed dolphin, where
intelligence is defined as adaptive flexibility in behavior. These
findings are in keeping with expectations based on the large size
of the dolphin brain, especially in neocortex development.
Dolphins extract information from their world through both vision
and hearing, the latter including a highly developed and
specialized echolocation (sonar) system. Studies have shown that
bottlenosed dolphins have an excellent memory for sounds heard and
things seen. Recent discoveries have shown that their sonar gives
them highly accurate 3-D information about their environment that
they can easily integrate with their visual world. Additionally,
dolphins have an excellent ability to learn rules and concepts, to
learn through observation, to imitate sounds or behaviors of
others, and to understand symbols as references to things or
events in the real world.
Scientists have long sought to find ways to enhance
communication with these animals. Attempts to decode their
vocalizations using different kinds of acoustic interfaces or even
to teach them to produce sounds that resemble English phrases were
not successful. On the other hand, dolphins can learn to
understand instructions conveyed by sequences of sounds or
gestural signals. In so doing, they take account of both the
meanings of individual symbols (the semantic component) and of
symbol sequences (the syntactic component). Moreover, they can
understand these sequences even when presented on a television
screen. Even degraded images can be interpreted.
We have been exploring the possibility to use video tracking
and other virtual reality technology to build a new interface
between humans and dolphins in a shared virtual space. This
approach is highly adaptable and allows for both analogue
(interaction with virtual objects) as well as symbolic (pointing
and clicking at symbols) interactions. Practical aspects of a
computer-based interface, such as protection of the electronics
against saltwater while still allowing unrestricted access for the
dolphins, have to be considered as well as communication with
remote human participants via the Internet. The video interface
will also allow new forms of interactions between dolphins and
computers (games) as well as between dolphins and their
environment. For instance, dolphins could turn on and select TV
and music programs, and operate other utilities in their
environments such as water faucets etc. With the capability of
computer based interfaces to store messages one can also expect
the emergence of new forms of dolphin-dolphin interactions, both
synchronous as well as asynchronous. While this interface can
provide new insights into dolphin communication, it can also
enrich the environment of captive dolphins. Once the basic
interfaces and communication protocols are established, wireless
communication networks will open new, exciting possibilities for
human-dolphin interactions.
Power-Law Distribution of the World Wide Web, Science
Scientist -especially those with background in physics-
working in the area of complex and non-linear dynamical systems
are constantly on the lookout for power-laws. That means that some
quantity changes as a function of another quantity not
proportionally but proportional to the quantity raised to a
certain power. This implies that details and specific scales of
the system are not important and therefore the property is
considered "universal" (some mathematicians joke by calling it "at
least inter-galactic"; the term means that the property holds for
a "class" of systems and one can therefore model it with its
simplest representative).
- Power-laws have been also observed in the WWW: Adamic et al.
counted how many websites are the targets of more than a given
number of other websites (i.e. have a given number of incoming
links) then this function displays a power law with exponent 1.94.
That means if 10% of the sites have one or two links pointing at
them then only 4.5% of the sites have up to 3 links pointing at
them, only half a percent would be the target of up to ten links.
The same property apparently also holds (with different
exponents) for the network of movie actors, the power grid etc.
The simple rule that gives rise to this universal behavior is
according to Barabasi et al.: (i) networks expand continuously by
the addition of new vertices, and (ii) new vertices attach
preferentially to sites that are already well connected. As a
consequence they predict that "old" websites tend to have more
connections than sites that are created more recently. This claim
was challenged by Adamic et al. based on statistics from the
InterNIC database and the argument that "sites that used to be
bland continue to be bland.
While scientists with physics background often come up with
novel ideas they are notorious in that they forget to do their
homework in statistics. Barabasi et al. point out that the age
related power law comes out nicely if one does the averaging
right.
Oh, yes, according to Einstein, physicists are also bad in
getting the sign right: Fig. 1A of Adamic et al. has the wrong
sign in the exponent of the distribution function. By the way,
that would lead to unbounded probabilities; small details for a
physicist.
MIT Prof Who Could Level The Cyber Playing Field, Businessweek
When it became clear that the Internet would create a new
way of globally networking people the connection to
self-organized, complex, adaptive systems was almost obvious. The
question was and still is about what types of new networked
structures would emerge. Some groups started talking about Global
Brains, others saw new opportunities for doing business. Whereas
in the first category one expects more a new form of intelligence
through the integration of networks of computers and humans in the
second category the objective is to design autonomous software
agents that exploit the information on the network.
Maes came up with an idea of "collaborative filtering" that
basically gives individual shoppers recommendations based on the
choice of other users. This method can be clearly useful in
finding "taste-mates" or situations like finding books that people
read who had read the book that I am just about to read. But it is
also clear that there is an instability for a convergence towards
a common mean (like in the choice of fast food or PC operating
systems). This is especially the case if the collaborative
filtering is exclusively based on frequencies in the sense that
only the most frequent choices are presented.
Maes latest economic enterprise is Open
Ratings that is supposed to rate web sites like hotels by
assigning them up to five stars based on the satisfaction levels
of customers. She hopes that this can help small stores to compete
with big name brands. It is clear that as soon as ratings become a
commodity there will be strategies to obtain those ratings at
market prices. And if a big company thinks it needs to drive out a
small company then perhaps buying the appropriate ratings would be
the cheapest way of doing that. According to Maes with the current
implementation you are not allowed to "say bad things about the
store where you have never shopped."
I am just waiting for the first rating companies to emerge to
sell any "open" ratings about any company. The expenses for the
obligatory shopping to become a customer would be included in the
bill. If we can learn anything from complex systems then it is
that there are no fool-proof strategies that work once and for
all. It is a nice idea, though.
Why Cisco Is The Atlas Of The Internet, Businessweek
It didn't take long in the first year of the new
millennium (or the last year of the old one) that the "plumbing
supplier" of the Internet surpasses the mother of all PC companies
in market capitalization. I remember very well the days when we
installed "Cisco boxes" when I was a post-doc at the Los Alamos
National laboratories in the mid-eighties. Nobody would have
imagined in their dream that this company (and not Radio Shack)
would one day be worth more than half a trillion US$.
In those days we were just about to replace dumb VT-100
terminals that we used to connect to Cray super-computers by smart
Macintosh computers with NCSA telnet terminal emulators. We also
would not have imagined that super-computers would one day be
replaced by networked workstations and personal computers. Today
we see more and more evidence that like in the brain the
connections are at the core of the performance.
With an exponentially increasing flow of information powerful
routing of this information becomes essential. It is hard to
predict how long Cisco and the technology it represents will
dominate the technology market. But it is clear that we are
nowhere near a plateau in terms of performance of Internet
traffic. Within a few years a large fraction of the traffic will
be routed on wireless networks. Economic, entertainment, and
increasingly environmental data will demand a bandwidth that will
continue to be a challenge for Cisco and competitors for a number
of years.
Links & Snippets
Do We Have A Knuckle-Walking Ancestor?, Nature
Bipedalism has been regarded as the fundamental
adaptation that sets hominids apart from other primates. Fossil
evidence demonstrates that by 4.1 million years ago, hominids
exhibited adaptations to bipedal walking. The fossil record offers
little information about the origin of bipedalism and the mode of
locomotion that preceded bipedalism. Richmond et al. present
evidence that fossils attributed to Australopithecus anamensis and
others retain specialized wrist morphology associated with
knuckle-walking. This suggests that bipedal hominids evolved from
knuckle-walking ancestors that were already partly terrestrial.
Sciences of the Interface, Announcement
On the occasion of Otto Rössler's 60th birthday
an international conference will take place in Karlsruhe in May of
2000. As is clear, the science of the interface is at the
epicenter of Rössler's research. Besides his pioneering
contributions in systems theory and chaos, he has concentrated his
recent work in the area he calls endophysics where the interface
between human mind and the rest of the world plays the central
role. Rössler's work has also inspired and provoked numerous
scientists and media theorists to reconsider the cartesian cut -
interfaces that can be either located or not.
- Sciences
of the Interface, An International Symposium on the
historical, Philosophical, Mathematical, Physical,
Biological, Social, Artistic, and Technical Aspects of
the Interface,
ZKM Karlsruhe,
May 18-21, 2000
3rd International Conf. on Complex Systems, Announcement
This is the third in a series of conferences with two
major aims: first, to investigate those properties or
characteristics that appear to be common to the very different
complex systems now under study; and second, to encourage cross
fertilization among the many disciplines involved.
PEDAGOGICAL SESSION: The conference will include pedagogical
sessions on Sunday, May. 21 covering fundamental knowledge in
concepts, simulation, and analysis tools relevant to complex
systems.
- Third International Conference
on Complex Systems, May 21-26, 2000, Nashua,
NH
-
Mastering Complexity -- Doing It Not Just Talking
About It -- Themes: E-Commerce, Knowledge Management, Health Care
"Managing the Complex" is a unique event -- 3 full days of
discussions and problem solving focused on the relationships
between managing organizations and the science of complex systems.
Our aim is to be learning together with a focus on solving
problems not merely intellectual chat.
