Comparing E. coli to Linux

[Nathan McCorkle]

This came out yesterday... awesome:

http://www.pnas.org/content/early/2010/04/28/0914771107

--
Nathan McCorkle
Rochester Institute of Technology
College of Science, Biotechnology/Bioinformatics

--
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To post to this group, send email to diy...@googlegroups.com.
To unsubscribe from this group, send email to diybio+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/diybio?hl=en.

[Bryan Bishop]

On Tue, May 4, 2010 at 1:36 PM, Nathan McCorkle wrote:
> http://www.pnas.org/content/early/2010/04/28/0914771107

Comparing genomes to computer operating systems in terms of the
topology and evolution of their regulatory control networks
pdf redirect: http://www.pnas.org/content/early/2010/04/28/0914771107.full.pdf+html

"The genome has often been called the operating system (OS) for a
living organism. A computer OS is described by a regulatory control
network termed the call graph, which is analogous to the
transcriptional regulatory network in a cell. To apply our firsthand
knowledge of the architecture of software systems to understand
cellular design principles, we present a comparison between the
transcriptional regulatory network of a well-studied bacterium
(Escherichia coli) and the call graph of a canonical OS (Linux) in
terms of topology and evolution. We show that both networks have a
fundamentally hierarchical layout, but there is a key difference: The
transcriptional regulatory network possesses a few global regulators
at the top and many targets at the bottom; conversely, the call graph
has many regulators controlling a small set of generic functions. This
top-heavy organization leads to highly overlapping functional modules
in the call graph, in contrast to the relatively independent modules
in the regulatory network. We further develop a way to measure
evolutionary rates comparably between the two networks and explain
this difference in terms of network evolution. The process of
biological evolution via random mutation and subsequent selection
tightly constrains the evolution of regulatory network hubs. The call
graph, however, exhibits rapid evolution of its highly connected
generic components, made possible by designers’ continual fine-tuning.
These findings stem from the design principles of the two systems:
robustness for biological systems and cost effectiveness (reuse) for
software systems. "

- Bryan
http://heybryan.org/
1 512 203 0507

[Andrew Hessel]

Thanks, Nathan, for bringing this to my attention.  This is a terrific paper.

On Tue, May 4, 2010 at 2:42 PM, letsoundragone <letsoun...@gmail.com> wrote:

really awesome!  thanks for sharing.
please send any article comparing genomes to computer operating systems

İlkim Tüzüner
Beyoğlu, İstanbul

2010/5/4 Nathan McCorkle <nmz...@gmail.com>:

[William Heath]

Finally the realization that synthetic biology may be a sub domain of computer science.

-Tim

P.S.

Long live the new computer science of synthetic biology!!! :>

[Anselm Levskaya]

The most interesting thing about the paper (and work like it) is the
comparison of the statistics of evolution in the context of natural
evolution vs. artificial evolution (of complex systems like linux).
They're pretty different kinds of evolution, utterly different objects
of reality, and facing totally different kinds of selective pressures.

Let's get one thing straight: a genome is not an operating system.
It's just a megabase or so of nucleotide letters. More like a
pickled/marshalled/serialized copy of the operating system. A
serialization with crazy encryption that nobody has any idea how to
crack yet.

The operating system is *Physics* acting on a set of several thousand
different polypeptide species floating around in a
lipid-membrane-delimited aqueous environment along with the
aforementioned polynucleotide genome.

As it turns out, *Physics* is a pain in the ass to reverse-engineer on
any irregular system bigger than a few atoms. You can't simulate
anything explicitly with QM at room-temperature, which means
everything's going to be pretty wrong unless you calibrate it to
measured data. (Hence the success of Rosetta vs. de-novo protein
simulation attempts.) I cannot emphasize enough how hard it is to go
from laws+actors --> predictions at these scales.

CS is centrally about abstracting away physical reality to enable
deductive reasoning in a synthetic, symbolic universe where we can
understand things. I understand the appeal of having biology live in
that universe. However...

Biology did not evolve to be understood. It evolved to work. It did
not evolve to have convenient hierarchies and tight ontologies, it did
not evolve to support abstraction layers. As such all abstraction
layers invented concerning biology tend to be leaky as hell and of
provisional value at best.

I can't tell you the number of bio-enthusiasts who've come from CS or
physics only to end up doing irrelevant theoretic/toy masturbatory
projects because they were too attached to their metaphors. If you
want to do biology, for now and for the foreseeable future, you are
going to need to be willing to get dirty and wet.

-a

[PYROcomp]

As my bio-physics professor used to say:

If it stinks it's organic chem,  if it's wet and slimy it's biology and if it just plain doesn't work it's physics...

       .___.
     /          \
    | O _ O |
    /    \_/    \
  .' /           \ `.
 / _|           |_ \
(_/ |           | \_)
      \         /
  __\_>-<_/__
     ~;/     \;~
Don't mess with the penguin!

[William Heath]

I can prove to you its a sub-domain of computer science.  Just because there are some unknowns about predictability, this is becoming less of an issue.  Once we hit a level of repeatability, biologists will wish for the days of non-predictability because computer scientists will be running the show.  Let me help you understand what computer science is all about:

The definition of a program:

A finite set of instructions to accomplish a goal.

I have been studying cells and comparing them to how silicon based computers work for years now.  I can accurately specify how they are actual computers now.  Brownian motion is the world wide web of the cell.  

DNA + cells are creating repeatable, stable, and complex working machines all the time.  What is the percentage of babies being born with 5 fingers?  It is soooo frustrating as a computer scientist to watch biologists flounder as they re-learn the discipline of computer science via biology with cells and DNA.  Synthetic biology is going to revolutionize the way computation occurs and radically change manufacturing.  What you hinted at, and what few people really understand, is that a cell is a quantum computer as well.  I actually believe it is even more than a quantum computer, that it is the window to even more advanced technologies and approaches.  I am completely dumbfounded by the resistance of biologists to what is completely obvious to almost any computer scientist.  You can't simply dismiss this with repeatability and current unstable outcomes via primitive lab approaches to dna program execution.  This will be solved and sooner than you think.  I would love to hear your analysis of what will happen when this stability issues are resolved.

-Tim

[AARON LEWIS DINKIN]

And let us not forget that viruses LITERALLY inject malicious code into "bacteria" how is that not like in the computer world? 

I'm really looking forward to see if we would be able to "cluster" or even "cloud compute" (with the theory of shared cellular memory) with cells as computers.

I'm wondering if "wifi" would even be tangible, we are technically electrical and respond to said stimuli in as of current "unknown" ways, who knows what we would be capable of in fifty years?

Reading BluRay?  Probably not.  Using Viri to rewrite dammaged DNA or genetic diseases YES!

[David Askirk]

On Wed, May 5, 2010 at 8:57 AM, William Heath <wgh...@gmail.com> wrote:
> I can prove to you its a sub-domain of computer science.  Just because there
> are some unknowns about predictability, this is becoming less of an issue.
>  Once we hit a level of repeatability, biologists will wish for the days of
> non-predictability because computer scientists will be running the show.

That is so true. I have a master in CS and just begun on bio. And I
think that it is easy to a certain degree because I can use my CS
skills. The hard part is learning the new words :-) But as a CS do I
love the way that everything is new!

--
Master Wizz. David Askirk
Hoeje Gladsaxe 64, 1. tv
2860 Soeborg
http://www.askirk.dk
Email: davse...@gmail.com
Skype: davsebamse
12884901877 er et primtal

[Anselm Levskaya]

@tim
1) Anyone who uses the word "prove" in the context of physical reality
already has more epistemological issues than a simple email is going
to solve.

2) If you're working with living cells or viruses or proteins for fun,
knowledge, or profit, who *gives a shit* about what you call what
you're doing? On the subject of categorization I commend the timeless
essay by Borges, conveniently hosted by a CS luminary:
http://www.crockford.com/wrrrld/wilkins.html

3) We'll get repeatability in biology by building a biology that's
repeatable, not before. In the meantime we'll make due with beating
the house.

4) You'd do well by knocking off the arrogance. There's only about a
million biologists in the world. You think you can beat the
computational power of that distributed hivemind? Learn from it
instead of giving it shit.

> A finite set of instructions to accomplish a goal.

Under this rubric you could shove the near entirety of human technical
knowledge. At which point the rubric becomes somewhat useless as a
conversational tool.

> now.  Brownian motion is the world wide web of the cell.

How the hell can you compare an *entirely random* interaction dynamic
with a precisely specified set of protocols and a near-fixed physical
network topology?

> DNA + cells are creating repeatable, stable, and complex working machines

These "stable complexes" exist in the framework of statistical
mechanics, not a deterministic state machine. It's certainly a
fascinating question how you design, build, and verify computational
systems that are -intrinsically- probabilistic. I would say given the
absolute dominance of the von neumann abstraction that we haven't yet
really treated the problem in great detail as of yet. The most
interesting work along these lines that I've seen lately has to do
with a natively probabilistic bayesian extension to functional
programming as detailed in the thesis by Vikash Mansinghka:
http://web.mit.edu/vkm/www/vkm-dissertation.pdf

> What is the percentage of babies being born with 5 fingers?

About 99.8% Polydactyly (extra fingers) occurs in 2 out of a thousand
births, roughly. Oligodactlyly (fewer than 5) is much more rare,
typically inherited.

> It is soooo frustrating as a computer scientist to watch biologists
> flounder as they re-learn the discipline of computer science via biology

It's soooo frustrating to watch formalists shoe-horn the world into
their preconceived models.

> and what few people really understand, is that a cell is a quantum computer
> as well.  I actually believe it is even more than a quantum computer, that

It's a quantum computer in the trivial sense that every physical
object is a quantum computer. But in the precise sense of encoding
variables into coherent superpositions, evolving the variables before
decohering into an output state that results in an algorithmic
operation, living cells do no such thing. They're slimy and hot, and
superpositions degrade in a femtosecond or less. Max Tegmark has done
careful calculations along these lines:
http://space.mit.edu/home/tegmark/brain.html

> completely dumbfounded by the resistance of biologists to what is completely
> obvious to almost any computer scientist.  You can't simply dismiss this

Nothing is obvious. Modern CS inherited a century of electronics
research that enabled the super VLSI that enabled powerful
instantiations of von neumann machines. It also inherited millennia
of mathematical, deductive reasoning. We're trying to
reverse-engineer four *billion* years of evolution. Do you really
think those aeons have no fundamental lessons for us? CS would murder
to understand how that which writes these words, writes these words.
It does not yet understand.

> with repeatability and current unstable outcomes via primitive lab
> approaches to dna program execution.  This will be solved and sooner than
> you think.  I would love to hear your analysis of what will happen when this

http://en.wikipedia.org/wiki/Hofstadter's_law

My life work is making biology fast and fun so that the deductive
tools of math&CS can be brought to bear on the problems. It will
happen, but it's going to take an ocean of sweat. I think that the
DIYBio movement could help a lot actually, if it takes the biology
seriously.

...And what will happen when we understand biology is that all hell
will break loose. Fun times!

[rwst]

Thanks to Anselm Levskaya for that Borges link.
There's a saying by Bismarck: "The 1st generation creates value, the
2nd manages it, the 3rd studies art history, the 4th degenerates." I
would like to state a version regarding computer science: "The 1st
generation creates algorithms, the 2nd writes and patches app
software, the 3rd merely knows how to download and install safely, the
4th blindly believes in computer power." OTOH, there have been people
who started at four and made it to position one, all in one person, so
don't take the saying as a prediction of the future, although
sometimes it looks like it.

How could one predict the climate of a planet you haven't seen the
geography of? Likewise, how do CS bio types expect to predict the
function of a protein where only the amino acid sequence is known?
They won't even get over the 90 per cent mark when it comes to
secondary structure. Mark my words. It's like a speech recognition
trying to spell a new word it has heard. Have you heard about the
tricky things even such a small molecule as water is able to perform
in vivo, when interacting with itself? There is the quantum at work.
You have to look at small things first to see how uncharted even
elementary chemistry is.

Evgenii S. Stoyanov, Irina V. Stoyanova, Christopher A. Reed: The
Structure of the Hydrogen Ion (Haq+) in Water. In: J. Am. Chem. Soc.
2010, 132, 5, S. 1484–1485, doi:10.1021/ja9101826

Abstract:
The hydrogen ion in water, Haq+, is a unique H13O6+ entity that
defines the boundary of positive-charge delocalization. Its central
unit is neither a C3v H3O+ Eigen-type ion nor a typical H5O2+ Zundel-
type ion. IR spectroscopy indicates that the H13O6+ ion has an
unexpectedly long central O···O separation (2.43 Å), showing that in
comparison with the gas and solid phases, the environment of liquid
water is uniquely proficient in delocalizing positive charge. These
results will change the description of Haq+ in textbooks of chemistry,
and a more extensive delocalization of positive charge may need to be
incorporated into descriptions of mechanisms of aqueous proton
transport.

http://pubs.acs.org/doi/abs/10.1021/ja9101826

Regards,
ralf

[OgnenSpie]

Hi all,

First time poster here, long time lurker ;)

On May 5, 10:49 am, AARON LEWIS DINKIN <cyan....@gmail.com> wrote:
>
> I'm really looking forward to see if we would be able to "cluster" or even
> "cloud compute" (with the theory of shared cellular memory) with cells as
> computers.

You might want to look into the areas of Membrane Computing and
Computing by Observing, where the meeting of theoretical computer
science and biology is producing interesting and wonderful
offspring...

Regards,
Ognen

[pRoSeCutin]

Dear All,

I am also posting for the first time like Ognin :-) Discovered this
community only yesterday. The PNAS paper cited here was interesting.
Personally I started my academics in the beginning of the last decade
as an electronics engg student for 3 yrs and switched over to
chemistry and finally found an interest in Biochemistry and molecular
biology. Being familiar with both Operating system concepts and the
DNA-Protein biosynthetic machinery, the similarity in their operation
became clearer with time. However my perception regarding the PNAS
paper is different, although it was not dictated through the paper,
and neither it was supposed to dictate, I would like to point to my
comments in:

http://www.nature.com/nature/journal/v465/n7297/full/465422a.html#comment-11308

Anselm Levskaya did a commendable job in putting his/her thoughts into
words. I am not going to technically dissect the previous postings
though. However, I would like to appeal with few propositions and
thought provoking questions with some very crude mathematical
estimates of my own:

The evolution of an OS is driven by different needs than the evolution
of cell. However we can understand much more by comparing these two
systems of evolution. Still there are other systems that continue to
evolve in similar and dissimilar ways: The languages and the
religions, basically any thing that has to do with artistic tastes or
a particular belief. Instead of just comparing evolution of OS and the
cell, we can understand lot more about the process of evolution, the
principles that drive change, by studying and comparing evolution in
other areas of knowledge. Consider one simple case: the evolution of
poetry... The human language is considerably much more complex than
either the permutations/combination possible in the realm of binary
system of a computer. I guess very few of us are familiar with the
study of evolution in languages as this is mostly a platform for
scientists, although we are familiar to some extent with poetry.... It
is so simple yet ordered structure, yet so diverse... each and every
poem is unique... If we try to think of the cell as a kind of
programmed device, what kind of program is there that can produce a
poem???

Same thing about dreams... yes dreams that we see at night.... dreams
are not very well analyzed by scientists because we can never provide
evidence of what we see. Even if we spend ~40% of our whole life
inside the world of dreams, we rarely analyze this data with
attention. I try to imagine the individual dreams to random cinema
(movie) that is being produced by an unknown entity, most of the times
very well-directed and well-enacted. How is it so well-structured?
Let us assume that dreams can be recorded by tapping wires into the
brain and saved in conventional optical/magnetic media........... I
guess we will need more bits to represent emotion and feelings, rather
than just frame by frame images.... Now the question comes: can we
represent all this data by corresponding arrangement of molecular/
cellular states inside the brain.... I guess theoretical calculations
can be done to check out the extremes.... Dreams are generally high-
res videos with 3-D perceptional levels, not just a 2-D depiction and
dreams are not generally well structured in time... e.g. we may have a
short nap, say 15 min but sometimes the dreams may contain so many
events within that 15 min which would take longer time when we are
wake up.... the basic observation is that dream can run faster or
slower in terms of time, but on average one night's dream might be
comparable to 9 hours of High-res 3-D video: say equivalent to 20 GB
of Blu-ray quality video (not sure, but for this discussion I assume
so). A human being lives for 80 yrs say. The total dreaming time is
40% of 80 year = 32 yrs. 32 yrs of High-res 3-D video is approximately
= 32x365x20 GB of data = 233,600 GB of data. How the brain, comprised
of cells comprised of molecules can create so huge amount of ordered
well-structured and meaningful data and process it and why?? Read
about evolution of behavior from "The Selfish Gene" by Richard Dawkins
and then try to figure out using the game theory approach why and how
human beings have evolved to watch dreams for ~40% of their entire
life??? Prof Dawkins have also discussed about evolution of memes...
please read that chapter before putting serious thoughts. A food for
thought for CS guys??? I have many more more personal thoughts in
this line of thinking.... you will find out a lot more surprises in
terms of how biological systems compute????

Very lazy to type... sorry not to pen down many more thoughts.... Want
to see some feedback from the group members...!!

Regards,

Indrajit

[Brian Degger]

thing I understood only recently, is that the methodologies of science
are as remixable as anything else.
For DIYBio I want minimal repitition, recyclable aparatus, no toxins
that need more than a bath room to dispose of.
My thoughts only,
Cheers
Brian

[Jonathan Street]

I agree that comparing biological systems to computing systems (at least as we currently recognise them) has limitations.

I'm struggling to follow your reasoning with regards to dreams though.  Firstly I'm not sure you can realistically compare a dream to high definition video.  Yes a dream is presented in the highest definition you can conceive with sound, smell, touch etc but all the imagery is already present.  You don't need additional space to store it just a database of previous events/people/places etc.  If you absolutely must force dreaming into a computer equivalent perhaps a video game would be a more realistic comparison.  You can play the game a thousand times over but everything is still coming from a single DVD.  Secondly I'm not clear as to why you think these dreams add up over a lifetime.  Most people struggle recalling their dreams at all.  You may recall a couple of favourite dreams from childhood but 99+% are lost.

The brain is certainly an amazing organ and for many tasks far surpasses what is currently possible in silicon but I don't think high fidelity data storage is one of those areas.

[pRoSeCutin]

I admit that my estimates about dreams are very crude. Yes, I already
thought of comparing dreams with video games. However, my point was
not to stress upon the memory aspect, but on innovation aspect of the
brain processes. Sometimes the dreams are repeated permutations of
familiar imagery, but for some people, and for some situation dreams
are very innovative and novel in nature. Dreams present a highly
challenging scenario for traditional science as nothing can be
presented as a verifiable and reproducible evidence with our current
repertoire of techniques. 99% of the time dreams are basically
repetitive, but leaving out those regular cases, there are some dreams
that are highly and exceptionally innovative.... I am talking about
some special ones which I am going to exemplify in little detail:

http://www.mindpowernews.com/BrilliantDreams.htm

some of these are not very well-known, but I have read about Kekule's
dream earlier in which he discovered the structure of benzene.

Personally, I used to see different poems in my dreams and sometimes I
could remember part of them and jot them down as soon as I wake up.
You may find many other historical examples where many other poets
have written down compositions that they saw in their dreams. Now my
question is: How can a brain do work and innovatively show us an
entire big poem with all complicated calculation of rhymes
unconsciously, which otherwise would take long time to figure out in
conscious wake-up state??? You may not believe, but I saw entire books
in my dreams that may not exist in real life and i could read them as
clear as I can read files on my laptop. Its a huge amount of data and
a result of huge amount of analytical thought that could result in
such a book that the unconscious part of the brain showed me in
dreams....It could take years to think and write such a book and the
brain creates it and shows it in just a moment without having any
precognition in our conscious part of the brain about what the
unconscious part is doing and why... It is not just comparable to one
video game dvd being played different times and producing little
variation all the time.... And why human beings have evolved to see
dreams at all as an organism? What kind of selective pressure was
there in the history of earth that compelled living organisms to
evolve to humans who see dreams 40% of their whole life? And a most
notable fact is that there is no such human being on earth who can
live without experiencing the dreams... it is as compulsory as eating
foods and drinking water.... why??? Why dreams formed an exclusive
requirement for living as a human??? And is there an animal version of
dreams that evolved into human version of dreams just like the
evolution of physiological features of other mammals has evolved into
the physiological features of humans...??? Many more questions need to
be asked before we start to think what could be the answers....

On Jun 24, 6:20 am, Jonathan Street <streetjonat...@gmail.com> wrote:
> I agree that comparing biological systems to computing systems (at least as
> we currently recognise them) has limitations.
>
> I'm struggling to follow your reasoning with regards to dreams though.
> Firstly I'm not sure you can realistically compare a dream to high
> definition video.  Yes a dream is presented in the highest definition you
> can conceive with sound, smell, touch etc but all the imagery is already
> present.  You don't need additional space to store it just a database of
> previous events/people/places etc.  If you absolutely must force dreaming
> into a computer equivalent perhaps a video game would be a more realistic
> comparison.  You can play the game a thousand times over but everything is
> still coming from a single DVD.  Secondly I'm not clear as to why you think
> these dreams add up over a lifetime.  Most people struggle recalling their
> dreams at all.  You may recall a couple of favourite dreams from childhood
> but 99+% are lost.
>
> The brain is certainly an amazing organ and for many tasks far surpasses
> what is currently possible in silicon but I don't think high fidelity data
> storage is one of those areas.
>

> On 22 June 2010 23:12, pRoSeCutin <indrac...@gmail.com> wrote:
>
> > Dear All,
>
> > I am also posting for the first time like Ognin :-) Discovered this
> > community only yesterday. The PNAS paper cited here was interesting.
> > Personally I started my academics in the beginning of the last decade
> > as an electronics engg student for 3 yrs and switched over to
> > chemistry and finally found an interest in Biochemistry and molecular
> > biology. Being familiar with both Operating system concepts and the
> > DNA-Protein biosynthetic machinery, the similarity in their operation
> > became clearer with time. However my perception regarding the PNAS
> > paper is different, although it was not dictated through the paper,
> > and neither it was supposed to dictate, I would like to point to my
> > comments in:
>

> >http://www.nature.com/nature/journal/v465/n7297/full/465422a.html#com...

> > diybio+un...@googlegroups.com<diybio%2Bunsu...@googlegroups.com>
> > .

> > > For more options, visit this group athttp://
> > groups.google.com/group/diybio?hl=en.
>

> > --
> > You received this message because you are subscribed to the Google Groups
> > "DIYbio" group.
> > To post to this group, send email to diy...@googlegroups.com.
> > To unsubscribe from this group, send email to

> > diybio+un...@googlegroups.com<diybio%2Bunsu...@googlegroups.com>
> > .