How to picture the cell?

atyy

Intelligence, to the extent that it is a useful concept(s) in science, never means that the laws of physics are violated. As an example, this link from the evolution faq uses the term "intelligence" in a way consistent with the underlying laws of "physics" (in that artificial world). http://boxcar2d.com/index.html

I don't know about the specific case in the Feinberg paper, but there is a notion of "swarm intelligence" applicable in some circumstances to bees. http://www.scholarpedia.org/article/Swarm_intelligence

Here is a free article that discusses some molecular details that may be relevant.
http://www.ncbi.nlm.nih.gov/pubmed/15572455
"Here, we report on the identification of a substance produced by adult forager honey bees, ethyl oleate, that acts as a chemical inhibitory factor to delay age at onset of foraging. Ethyl oleate is synthesized de novo and is present in highest concentrations in the bee's crop. These results suggest that worker behavioral maturation is modulated via trophallaxis, a form of food exchange that also serves as a prominent communication channel in insect societies. Our findings provide critical validation for a model of self-organization explaining how bees are able to respond to fragmentary information with actions that are appropriate to the state of the whole colony."

Darwin123

And he is pointing out that it is your analogy, having nothing to do with any response others have made to your OP.
Leggo blocks are interesting in that they have been designed to have multiple connection points so that there is no one stable configuration to them. Oddly, this is the opposite of what you are saying. The shapes were carefully designed so that one can make more than one shape. In this case, the corporate chaos in their configuration is the result of design. Leggo's were designed to be random.
Not much different from the computer simulations that impressed you as being random. You did not see pictures of actual molecules forming structures with random motions. These were make believe molecules that were following random paths. The thing is that there could not be any mystical purpose to their motion because these cartoon molecules were programmed not to have any order in their motion.
You initially said that there had to be some order in the motion of the molecules that was the result of some programming or purpose. However, the motion that you saw had actually been programmed not to show any order. Random number generators were used, but the final results are independent of the random number generator chosen. There was no specific random number generator necessary to make the order emerge.
The order that you saw emerge from the cartoon molecules was not due to any organized motion. It came from their initial shapes. I conjecture that you are now going to claim that the shapes had to be programmed according to a certain order. However, that is an issue separate from the initial one of this thread. However, I did try to address that second issue already by referring to a study by Lenski and Blount.
The specific shapes are also due to another type of random variation in the heredity of the organism. Physically, the variation in heredity is random. Scientists have studied the physical processes of inherited variations such as mutation. Again, the direction of inherited variation is for the most part random. They arise through the random motion of molecules, that were already discussed. There is no programming in mutation. So how come the results of many accumulated variations show order? For instance, why do the molecules and processes show shapes far more specific than any Leggo block?
The shapes are determined by a highly unrandom process called natural selection. It is natural selection that makes a generalized shape into a very specific one.
I posted a link about an article by Lenski and Blount. This article showed how random mutations accumulated into very specific chains of chemical reactions. The shapes of some of the molecules changed to make certain chemical reactions more probable and other chemical reactions less probable.
The experiment was very tightly controlled so that each mutation could be tracked back to the original ancestor to whom that mutation occurred. For instance, the bacteria in their experiment were castrated! The cells were mutilated so that they couldn't perform any process similar to sexual reproduction. All these bacteria could do is split. They had to reproduce asexually.
Note that this would have seriously restricted the amount of variation that was possible. Castration should have prevented the bacteria from reaching their full evolutionary potential. The changes observed in this experiment were less than what would have been seen in "wild type" bacteria.
What were seen were huge changes in the chemistry of this bacteria. The castrated bacteria, among other things, evolved to consume citrate in an oxygen atmosphere.
There would probably have been even more amazing changes for a viral bacteria. More sex, more recombination, more inherited variation, and more evolution. So why were the bacteria castrated? To prove something that you are trying hard not to believe.
By tracking each mutation back to the source, Lenski and Blount proved that the mutations were random. No mutation occurred that was so complex it required a programmer to design. These were all "small" mutations. No miracle occurred in any one generation. They could prove it because they could track the changes.
So here is an example where natural selection, acting on "random mutations", resulted in an "improbable" change. The individual changes were random. However, the accumulated change was not random.
One other thing that may help you understand is the "organized" shape of a snowflake.
Usually we see snowflakes that are clumped together. So few people, at least the urban ones, have seen snowflakes just after they were formed before collision. However, I have been in the country in areas where the snowflakes fall without interference. They are intricate six-fold shapes with amazing order to them.
A few snowflakes are identical. Maybe as many as two snowflakes in a thousand are identical. However, this is a small fraction. The vast majority of snowflakes are not alike. They have amazing variety, yet each type of snowflake shows unmistakable order. Here is something you should consider.
The shape of a snowflake has nothing to do with programming either. A snowflake forms by diffusion. The water molecules in vapor are moving randomly. The water molecules, although simple, have specific shapes. However, each water molecule has a trajectory and an orientation that is random.
The order comes from another type of natural selection. The sticking probability of a snowflake is biased by heat exchange with the atmosphere. The water molecules tend to stick at points on the flake, not flat areas. So a region with lots of points gets more water vapor to make even more points. In other words, the points reproduce with survival of the fittest.
The variation in snowflakes is caused by random variation modulated by a selection process. It isn't selection in the sense of an organism. The selection does not have purpose as animals understand it. However, it is selection. The selection shapes the random process into an order pattern.
Diffusion does result in order. Small variations can accumulate with selection into an ordered pattern. That is self assembly.

Darwin123

I can do one better. I saw a demonstration just yesterday of plastic pieces, not directly controlled by anything, self assemble into a sphere. Here are links to similar demonstrations. You can watch the video and tell us who is controlling the switchboard of these plastic pieces.
The links have a video showing the self assembly of a model of the polio virus. Olson used a computer to carve models of pieces of the polio virus capsid. Magnets were placed at the edges of the pieces. The pieces fit together to form a spherical shaped body, which is a model of the entire polio virus.
I just attended a seminar by Olson where he demonstrated this “self assembly” among other things. After I the seminar, he let me shake the container myself. The sphere rearranged several times when it was “softly” shaken. Note that one can see many failed “attempts” by the pieces to rearrange themselves. However, these intermediate forms fall apart. The final result is one of two stable forms for the polio capsid.
Olson places the sphere in a sealed container. When he shakes vigorously, the sphere breaks up into separate pieces. When he shakes gently, the pieces assemble into the sphere.
One caveat. Without a core, there are two stable forms of the capsid. One stable form has a trapped piece inside the nearly assembled capsid. There is a hole in the capsid. Olson assured me that could be prevented by placing a model of the RNA core in the container. That is closer to the situation in a real cell, where the stable form with the hole doesn’t occur. However, he didn’t include the core in the container that I shook. the version of the simulation that I took part in had two outcomes. A capsid with a hole in it occurred about half the trials.
The motion is not “programmed”, beyond the amplitude of the shaking. The vigor of the shaking is analogous to temperature. However, the motion of the pieces is random. This is how the OP originally framed her question. The statement is there had to be some will governing the motion of the pieces. This clearly is not true.
If there is programming involved, then it is in the shapes of the separate pieces. There is no purpose and no will to the individual motions.
He also described experiment where he mixed two types of pieces of opposite chirality. When shaken, the mixture formed two polio virus models of opposite chirality. So the fact there is more than one “stable structure” is irrelevant. The random motion ends up separating the assembled pieces. I didn’t see that demonstration, but I have no difficulty believing it.
He also claims to have done the experiment with an automatic shaker. The pieces still assembled. There was no programming in the shaker. When the shaking was restricted to a range of vigor, the sphere self assembles.

Short link showing just the self assembly of the polio virus model.
http://www.youtube.com/watch?v=fKGulHzykBU
”Arthur Olson (www.mgl.scripps.edu) gives a very engaging demonstration of how molecular recognition and self-assembly can be explored using an innovative, hybrid user interface that combines 3D solid printing…”

Longer video with a more complete lecture. Basically, he is talking about a new type of computer interface which he used in making these pieces of polio virus. However, this discussion may give you a better idea of how biochemists see the universe.
http://vimeo.com/26198903
“Arthur Olson (mgl.scripps.edu) gives a very engaging demonstration of how molecular recognition and self-assembly can be explored using an innovative, hybrid user interface that combines 3D solid printing and his own augmented reality environment. He argues that a synthesis of both abstract & 3D visualization is needed to understand biological processes. He then presents impressive, cutting-edge methods for fast, integrated, and large-scale macromolecular visualization. This talk was presented at VIZBI 2011, an international conference series on visualizing biological data (vizbi.org) funded by NIH & EMBO. This video was filmed and distributed with permission under a creative common license.”