Alphabetical Site Index

David W. Roscoe

This Web Page is about me, David W. Roscoe.

It is about what I am, what I have done, what I am doing, what I hope to do, and how to contact me.

[Click on (Back) button now.]

My Formal Education.

In my early years in grade school, science and math were my favorite subjects. I liked them because I could understand them. They helped me to understand the universe. They also made me realize that there were amazing things, such as the landing of a man on the Moon while I was in grade school, that would be possible after we developed the technology based on science and math.

I did not like languages, history, social studies, and similar subjects. I could not use them to help me understand the universe because I saw little or no cause and affect in their subject matter. Unlike science and math, they often seemed to be matters of opinion instead of matters of fact. For me, studying them was boring memorization. And I could not see much practical use for them.

Later, in college, I studied electrical engineering. I specialized in computers. I liked computers because, when programmed correctly, computers are very predictable. Also I saw computers as tools that were potentially very useful in other activities.

My Career.

After college I became a software engineer. That is the fancy name for computer programmer. When I needed to earn some money, that is what I did. At first I worked as a salaried employee. Later I worked as a contractor.

In the beginning my interests were mainly algorithms and data structures. Later I became interested in networks, protocols, cryptography, and security.

When I changed jobs, I tried to find a job in which I could learn about the things in which I was interested. But that was not always possible. Employers usually want the people that they hire to already know all about the work that they want done. So sometimes I would go for long periods, living on savings, and keeping busy working at home on interesting projects of my own choosing.

I am in one of those periods now. I am very involved in My Pursuit Of Justice, especially Project Media Matrix and related matters.

But I am looking for paying work also. I am not spending a lot of time on it, because I have a comfortable amount of savings, but I know that the savings will not last forever.

My Resume

My Self-Education.

Eventually I became interested in social issues. These were the subjects that I avoided in school because they seemed complicated. But now I realize that they are not complicated. They simply were not taught very well in school, at least in the schools that I attended.

For example, I did not understand economics until I read the book Free to Choose : A Personal Statement by Milton Friedman, Rose D. Friedman (Contributor) though I had taken a course in economics earlier at college.

After I realized that social issues could be understood, I began to teach them to myself. But it was a slow process. Good reading material about social and political issues was much harder to find than material about science and technology. But I made progress.

My Pursuit Of Justice.

In grade school I was often a target of bullies. So I was more aware than most of injustices. I noticed that often bad things happen to good people and good things happen to bad people. As I became older it bothered me more, and I began doing something about it.

I gave contributions to charities that said that they would use the contributions to help the good people that were hurting. I gave contributions to political organizations that said that they would use them to end injustices by changing laws. But I became concerned that my contributions were being wasted. Many charities help people that are hurting, but do nothing to prevent new hurting of different people in the future. It it as if they are treating symptoms without treating the disease. And many political organizations end an injustice to one group of people, but do an injustice to a different group of people.

I became more careful of the causes to which I contributed, and began to contribute mostly to libertarian causes. The libertarian movement concentrates on reducing the size and scope of government, which people in the movement believe is the cause of the most serious problems in the world, but does not give preferential treatment to some groups at the expense of others. To learn about libertarianism and the organizations that promote it click on http://www.libertarianism.com/ .

But I was not satisfied with the progress that any organization was making at ending injustice in the world. I concluded that a major cause of the poor progress was the control of the flow of news about injustices, and I began Project Media Matrix to solve this problem. It now occupies most of my time.

Project Media Matrix [click to view]

Exercise Routine

For years I have had an exercise routine, at least jogging, to keep myself healthy.

This was my routine immediately before I noticed My Psychotronic Harassment.

If I was able to do a full 3 sets of 10 repetitions of an exercise with weight, then I would increase the weight slightly. The back pack was up to 48 pounds, and the free weights were up to 41 pounds in each hand, when My Psychotronic Harassment forced me to suspend my exercise routine at the age of 48.

I resumed the routine a few weeks later, but without the weights because I felt so poorly. Push ups and body lifts without the pack felt almost as difficult as they had been with the pack.

I eventually concluded that the difficultly was caused by Weakness With Psychotronics. I began using the pack weight again. I was able to do the exercises with the weight with approximately the same difficulty, but the number of repetitions that I could do was decreasing. The body lifts were the worst, which became as low as 1 or 2. I was lifting only 31 pounds of free weights in each hand.

In 2003 March I suspended the weight lifting. I now do only: stretching; and 30 minutes of running, outside if weather permits. I hope to resume the weight lifting eventually.

Contact Me

You should be able to Contact Me, David W. Roscoe, in any of the following ways.

If I do not know who you are, then please write something that interests me in the Subject field so I will know that your message is not more junk e-mail (SPAM). Please be specific. For example, write "Project Media Matrix Web Site" instead of "your web site". I ask this because in 2003 Jul, after I began ignoring the obvious SPAM, it began being disguised as persomal e-mail from strangers. If you do not do this then I might conclude that your message is SPAM and ignore it.

  • By postal mail to

      30 Worthen St Apt A9, Chelmsford MA 01824-2630

  • By telephone, if I gave to you my telephone number.

  • In person at the Freedom Forum of Nashua NH.

    I am the thin guy who, depending on weather and other conditions, is usually wearing some combination of the following:

    The Bad Guys sometimes disrupt my communications because of my work on Project Media Matrix. So...

    If you have difficulty contacting me, then please keep trying. The Bad Guys will not block us forever, because they do not want their interference to be very obvious.

    Hiring Me [click to view]

    Sending Money [click to view]

    [Click on (Back) button now.]

    Solution To The Protein Folding Problem

    In the summer of 2001, I developed what I believe is a simple computer solution to a problem in biochemistry known as the Protein Folding Problem (Live Web Page: here). But something strange happened when I began telling people about my solution.

    What is the Protein Folding Problem? It is to calculate the correct shape of a protein based on the DNA sequence of the gene that makes it. People say that a solution to it is worth a Nobel Prize because it will dramaticly increase our understanding of the living cell and diseases, and make the design of new drugs much easier.

    Your genes are chains of DNA units. These are translated into chains of RNA units. These are translated by ribosomes into chains of amino acid units. And these amino acid chains bend and fold up into protein molecules.

    The translations from one type of chain to another are well understood. What is not understood is how the amino acid chain folds up into a protein of a particular shape. There are many ways to fold a particular amino acid chain into a protein, producing many different shapes. But only one shape is correct, the shape produced in the living cell.

    Scientists have been trying simulate the folding process for decades using computers. Their problem is that they have been simulating the wrong folding process. Their process assumes that folding begins with the whole amino acid chain stretched out straight, and that it will fold itself into the correct shape. This works for some chains, mostly short ones, but does not work for all of them. And it is not what happens in the living cell.

    My solution simply simulates more accurately what happens in the living cell.

    Sequential Protein Folding
    by David W. Roscoe
    2001 Sep 04

    Sequential protein folding is a method of folding proteins and is based on the following suppositions:

    1. The ribosome outputs amino acid chains one amino acid at a time.

    2. A partial amino acid chain begins to fold before the entire chain is output.

    3. Protein folding consists of a sequence of smaller folding steps, each operating on a combination of a previously folded partial amino acid chain and the next amino acid.

    The pseudocode for the sequential protein folding algorithm is: 

      Protein = EMPTY;
  while not Complete(Protein)
    {
       AddNextAminoAcidTo(Protein);
       ApproachEnergyLocalMinimumOf(Protein);
    }

    In the simplest case the procedure ApproachEnergyLocalMinimumOf() actually finds the local energy minimum state of the amino acid chain. This is based on the suppositions that:

    1. There is sufficient time before the addition of the next amino acid for the chain to move to the local energy minimum state.

    2. There are no other interactions that affect the folding process, such as interactions between the amino acid chain and the ribosome, the messenger RNA, or other parts of the assembly of which the amino acid chain is a part.

    But the above suppositions might not be true, especially in cases of large proteins. To insure correct folding in all cases the procedure ApproachEnergyLocalMinimumOf() might need to account for factors such as the speed of the ribosome, and the physical and chemical attributes of other assembly components.

    It seems that sequential protein folding should provide better results than other methods about which I have read. But I earn my living as a software engineer. I do not have the time or other resources to do proper testing of sequential protein folding. I hope that I soon can read the results of tests done by others.

    I published this solution by e-mail to many people interested in similar technology, including the e-list of the Nanotechnology Study Group which meets in Cambridge, Massachusetts. And I attended one of their meetings and presented the idea.

    There was some interest, but not as much as I expected. And many people seemed to have difficulty understanding it.

    I did some research on the Internet and discovered that experimental scientists had confirmed what I suspected. Protein folding was co-translational, meaning that folding happens while the amino acid chain is coming out of the ribosome, and this effects the final protein shape. So anybody that tries to simulate protein folding starting with a whole elongated chain is completely wasting his time.

    To illustrate why this is true, consider the simple curling protein model below. The "o"s represent amino acids, of which there only one type. They can fold only by curling one way with 90 degree bends.

    Here is a stretched out chain of 18 units. 

     o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o
    Here is what would happen if the straight chain was released. It would fold into 2 back-to-back spirals. 
     o-o-o o-o-o
 |   | |   |
 o o-o o-o o
 |         |
 o-o-o-o-o-o
    But here is what would happen if the same chain was constructed by a ribosome adding one unit at a time from the left end and allowing it to fold while being constructed. It would form a single larger spiral. 
             o-o-o-o
         |     |
         o o-o o
         | | | |
         o o o o
         |   | |
         o-o-o o
               |
 (RIBOSOME): o-o
    This is obviously a different shape and illustrates why the computational chemists that are trying to produce the natural fold of real proteins from elongated chains are going nowhere fast.

    But they continue to try. In fact one group, called Folding@Home (Live Web Page: here), has organized people with thousands of Internet connected computers to work together to fold proteins using this method. These guys are going nowhere faster.

    Relatively few natural proteins, mostly the short ones, fold themselves. This is true of our curling proteins also. A 4-unit chain folds into a U-shape, with or without a ribosome. 

     o o                  o-o
 | |                    | U-shape is on it's side
 o-o      (RIBOSOME): o-o
    But in general proteins do NOT fold themselves.

    I found only one example of computer simulation of co-translational protein folding on the Internet at the Virtual Ribosome (Live Web Page: here), but it was incomplete.

    Here are some other factors known to effect folding that must be included in an accurate computer simulation of protein folding.

    Instead of the researchers making their model more similar to nature so it will fold more proteins correctly, they continue to use their existing model to experiment on only the few small proteins that fold with it. Why are researchers working so hard simulating a model of protein folding that is known to be inaccurate? It is puzzling. I have asked some researchers this question, but I do not understand the answers.

    I am becoming convinced that important scientific ideas such as the Solution To The Protein Folding Problem are being suppressed, maybe because an influencial drug company or military biowarfare department is using it for research. In addition to control of research funding, I suspect that the methods of this suppression include Mind Control With Psychotronics, so the researchers themselves might not realize it.

    The following is correspondence that I had with protein folding researcher Alan Fersht about protein folding, which also appeared on the Web (Live Web Page: here). 

    Date: Tue, 26 Nov 2002 23:44:13 -0500
To: ...
From: "David W. Roscoe" ...
Subject: NSG-D/ Protein Folding Mental Block, response
Cc: 
Sender: nsg-d-approval@world.std.com
Reply-To: nsg-d@europe.std.com

And then "David W. Roscoe" ... says:
Researcher Alan Fersht replied to my original questions about protein folding,
and we exchanged several messages.  Those messages are appended.


Dave.

----------------------

Date: Fri, 15 Nov 2002 22:57:01 +0000
To: "David W. Roscoe" <...>
From: Alan Fersht <...>
Subject: Re: Protein Folding Mental Block
Cc: ...

Dave

Some answers to your questions.

>Protein Folding Mental Block
>
>
>All the computer simulations of protein folding about which I have read
>begin the simulation with a complete pre-existing amino acid chain.
>But this contradicts what biochemists have known for some time about
>protein folding in the cell.  For example:
>
>(1)  Cotranslational folding, the folding of the amino acid chain as it
>emerges from the ribosome, has been proven.

Most theoretical and advance experimental work is done on very small
proteins that are single domains.  These generally do not fold until
all or nearly all of their sequence is present - see several of my
papers on CI2 and barnase, co-authored by Gonzalo de Prat Gay or
Jose-Luis Neira.  Accordingly, these cannot fold co-translationally,
but can only fold after they have been fully synthesized.  Further,
they probably can't fold until they have left the channel of the
ribosome.  So a complete pre-exisitn chain is a good model.


>
>(2)  The end of the chain to emerge first is attached to the surface of
>the ribosome during cotranslation folding, so most proteins have no
knots.

There is, I believe, only one protein that appears to have a knot.

>
>
>(3)  The use of redudant DNA and MRNA codons for some amino acids cause
>temporary but significant genome controlled reductions in ribosome
>speed at the boundaries between protein domains, apparently to allow
>time for a newly folded domain to settle with previous ones, before the
>output and folding of the next domain begins.
>
>These facts indicate that folding in the cell happens at ribosome
>speed, in a series of small fast steps, involves significant movement
>of only a few joints per step, and the natural fold is in general *not*
>the global energy minimum, in fact it can be far from it.  Inclusion of
>these factors in a folding simulation would dramaticly reduce
>computation requirements.  Such a simulation could probably fold almost
>any protein [those not needing chaperones] correctly in a few minutes
>on an ordinary desktop PC.  And the protein folding problem would be
>solved!

It is difficult enough, if not impossible in general, at present to
fold by simulation a small single domain protein de novo so I find it
difficult to understand how a larger protein can be solved now in a
few minutes on a PC.


>
>But computer simulations that ignore these factors obviously can not
>work except on the subset of shorter proteins and fragments that can
>also fold themselves in a test tube.  So why do all the protein folding
>computer simulations seem to be ignoring these major factors?  Why are
>researchers concentrating on minor factors such as energy functions, or
>trying to figure out how to pick out the natural fold from all the
>"decoys" produced by their simulations?  Why are they simulating an in
>vitro process instead of the known in vivo process?  Is the correct
>solution being suppressed for political reasons?

The in vivo processes are just far too difficult at present and so
researchers are concentrating on identifying features responsible for
folding in tractable systems.  Believe me, if any of the
theoreticians had solved the problem, they would be shouting it from
the rooftops and booking a flight to Stockholm in December.
Alan
--
Alan Fersht
Herchel Smith Professor of            		Director
Organic Chemistry				Centre for Protein Engineering
Department of Chemistry,			MRC Centre
Lensfield Road, Cambridge CB2 1EW		Hills Road, Cambridge CB2 2QH
...

------------------------

Date: Thu, 21 Nov 2002 01:14:32 -0500
To: Alan Fersht <...>
From: "David W. Roscoe" <...>
Subject: Re: Protein Folding Mental Block
Cc: "David W. Roscoe" <...>

Hello Alan.


At 10:57 PM 11/15/02 +0000, Alan Fersht wrote:
>Dave
>
>Some answers to your questions.

Thank you.  And I am sorry for taking so long to reply to you.


>>All the computer simulations of protein folding about which I have read
>>begin the simulation with a complete pre-existing amino acid chain.
>>But this contradicts what biochemists have known for some time about
>>protein folding in the cell.  For example:
>>
>>(1)  Cotranslational folding, the folding of the amino acid chain as it
>>emerges from the ribosome, has been proven.
>
>Most theoretical and advance experimental work is done on very small 
>proteins that are single domains.  These generally do not fold until 
>all or nearly all of their sequence is present - see several of my 
>papers on CI2 and barnase, co-authored by Gonzalo de Prat Gay or 
>Jose-Luis Neira.  Accordingly, these cannot fold co-translationally, 

I read some of your papers, including one of the papers that you mentioned
above, "Complementation of peptide fragments of the single domain protein
chymotrypsin inhibitor 2", at
http://www.mrc-cpe.cam.ac.uk/publications/ladurner_jmb_frag.pdf , in which you
found that the partial CI2 chains would not fold from the extended state.  But
these were in vitro experiments.

Do you think that they might fold if they received folding guidance from the
ribosome in vivo, such as the pushes and pulls caused by the cycling of the
ribosome, and the physical constraint of anchoring the N-terminus of the chain
near the exit port during translation?  If not then do you think that such
guidance might at least cause folding to happen faster after the chain is
complete?

It might be true that some domains and small proteins do not fold co-
translationally.  But apparently most other proteins do.  Would you agree that
a correct simulator, one that produces only natural folds, probably should
take this factor into account?


Maybe I should ask a different question.

I am a software engineer, and it seems to me that changing an existing folding
simulator to include the known in vivo factors would not be very difficult.
Add amino acids here, apply forces there, constrain movement there, and it
would be done.  These changes would not significantly increase the amount of
computation needed per time step.  But they might dramaticly reduce the number
of time steps needed for folding.  There is so little to lose, but so much to
gain.

So why has nobody done it?  The only example I found of anybody trying it was
on a page called the Virtual Ribosome at http://www.yasara.com/ribosome.htm ,
but it was incomplete and it's creator was busy with other things.


>It is difficult enough, if not impossible in general, at present to
>fold by simulation a small single domain protein de novo so I find it
>difficult to understand how a larger protein can be solved now in a
>few minutes on a PC.

An in vivo folding simulator might be faster for these reasons: (1) it needs
to simulate fewer and shorter folding pathways; and (2) it could simulate more
time steps per computer CPU second.

(1) fewer and shorter folding pathways:

Interdomain translational pause indicates that in vivo folding happens much
faster, in general, than in vitro folding, at close to ribosome speed, in time
approximately proportional to chain length.  My guess is that it is faster
because the in vivo factors eliminate all folding pathways except the one[s]
leading to the natural fold.  These in vivo factors would make a simulator
faster for the same reason.

(2) more time steps per computer CPU second:

The amino acids units in the emerging chain can be divided into 2 parts, an
unfolded part and a folded part.

The unfolded part consists of units that have not yet folded to a stable
state.  They are mostly units output recently by the ribosome.

The folded part consists of units that have already folded with other units to
a stable state.  Most of them were output earlier.  Except for vibration,
there is very little relative motion in the folded part.  The folded part can
lose units to the unfolded part, but mostly it gains units until it includes
all of the chain when folding is complete.

The unfolded part must be simulated in the usual way.  But the usually much
larger folded part can be mostly ignored by the simulator, maybe treated
simply as shaped masses, except for a few of its units near the unfolded part
or the ribosome.  So the simulator has less to do per time step, and can
simulate more time steps per CPU second.

Does this seem reasonable to you?


>Believe me, if any of the
>theoreticians had solved the problem, they would be shouting it from
>the rooftops and booking a flight to Stockholm in December.
>Alan

I believe you.  My problem is understanding why nobody has tried this idea yet.


I thank you again for your time.  I hope that I am not making a pest of myself.


Dave.

-
...

---------------------

Date: Thu, 21 Nov 2002 10:52:44 +0000
To: "David W. Roscoe" <...>
From: Alan Fersht <...>
Subject: Re: Protein Folding Mental Block

David

>
>Do you think that they might fold if they received folding guidance
>from the ribosome in vivo, such as the pushes and pulls caused by
>the cycling of the ribosome, and the physical constraint of
>anchoring the N-terminus of the chain near the exit port during
>translation?  If not then do you think that such guidance might at
>least cause folding to happen faster after the chain is complete?

No.  The proteins don't fold because they are thermodynamically
unstable when only partly formed.  We haven't a clue how the ribosome
could "guide folding", apart from vague notions of pausing at rare
codons.

>
>It might be true that some domains and small proteins do not fold
>co-translationally.  But apparently most other proteins do.  Would
>you agree that a correct simulator, one that produces only natural
>folds, probably should take this factor into account?


Possibly, If we knew how to simulate it.

>
>
>Maybe I should ask a different question.
>
>I am a software engineer, and it seems to me that changing an
>existing folding simulator to include the known in vivo factors
>would not be very difficult.  Add amino acids here, apply forces
>there, constrain movement there, and it would be done.  These
>changes would not significantly increase the amount of computation
>needed per time step.  But they might dramaticly reduce the number
>of time steps needed for folding.  There is so little to lose, but
>so much to gain.
>
>So why has nobody done it?  The only example I found of anybody
>trying it was on a page called the Virtual Ribosome at
>http://www.yasara.com/ribosome.htm , but it was incomplete and it's
>creator was busy with other things.

The theoreticians can't even cope with small domains that don't need
any help to fold!  Your suggestion is even more difficult.


>
>>It is difficult enough, if not impossible in general, at present to
>>fold by simulation a small single domain protein de novo so I find it
>>difficult to understand how a larger protein can be solved now in a
>>few minutes on a PC.
>
>An in vivo folding simulator might be faster for these reasons: (1)
>it needs to simulate fewer and shorter folding pathways; and (2) it
>could simulate more time steps per computer CPU second.

To reiterate, a simulator applying methods to speed up in silico
folding of large proteins would not be any faster than simulating
small proteins so it is not yet feasible.  But, it will be necessary
I suppose for complex proteins.


>
>(1) fewer and shorter folding pathways:
>
>Interdomain translational pause indicates that in vivo folding
>happens much faster, in general, than in vitro folding, at close to
>ribosome speed, in time approximately proportional to chain length.
>My guess is that it is faster because the in vivo factors eliminate
>all folding pathways except the one[s] leading to the natural fold.
>These in vivo factors would make a simulator faster for the same
>reason.

We have discovered small proteins that can fold in microseconds, and
these are just on the borderline of being analysed by current MD
simulation.  The time scale of in vivo folding is 6 orders of
magnitude slower.

>
>(2) more time steps per computer CPU second:
>
>The amino acids units in the emerging chain can be divided into 2
>parts, an unfolded part and a folded part.
>
>The unfolded part consists of units that have not yet folded to a
>stable state.  They are mostly units output recently by the ribosome.
>
>The folded part consists of units that have already folded with
>other units to a stable state.  Most of them were output earlier.
>Except for vibration, there is very little relative motion in the
>folded part.  The folded part can lose units to the unfolded part,
>but mostly it gains units until it includes all of the chain when
>folding is complete.
>
>The unfolded part must be simulated in the usual way.  But the
>usually much larger folded part can be mostly ignored by the
>simulator, maybe treated simply as shaped masses, except for a few
>of its units near the unfolded part or the ribosome.  So the
>simulator has less to do per time step, and can simulate more time
>steps per CPU second.
>
>Does this seem reasonable to you?
>
Small domains fold in a concerted manner, with the whole chain
involved.  If the folding of larger proteins can be broken up into
sequential folding of constituent domains, then calculation would
become orders of magnitude more efficient.

To summarise what I am saying is that in principle we do need to
simulate how the ribosome works in order to understand how larger
proteins fold in vivo, which may lead to a better understanding of in
vitro process and could speed up in vitro simulation.  And your ideas
may well be useful for these complex systems. But, we are years away
in computing power.
Alan
--
Alan Fersht
Herchel Smith Professor of            		Director
Organic Chemistry				MRC Centre for
Protein Engineering
Department of Chemistry,			MRC Centre
Lensfield Road, Cambridge CB2 1EW		Hills Road, Cambridge CB2 2QH
...

-------------------------

Date: Sun, 24 Nov 2002 23:44:50 -0500
To: Alan Fersht <...>
From: "David W. Roscoe" <...>
Subject: Re: Protein Folding Mental Block
Cc: "David W. Roscoe" <...>

Hello Alan.


Again I am sorry for taking so long to respond.  You gave to me much
about which to think and research.


At 10:52 AM 11/21/02 +0000, Alan Fersht wrote:
>David
>
>>
>>Do you think that they might fold if they received folding guidance
>>from the ribosome in vivo, such as the pushes and pulls caused by
>>the cycling of the ribosome, and the physical constraint of
>>anchoring the N-terminus of the chain near the exit port during
>>translation?  If not then do you think that such guidance might at
>>least cause folding to happen faster after the chain is complete?
>
>No.  The proteins don't fold because they are thermodynamically
>unstable when only partly formed.  We haven't a clue how the ribosome
>could "guide folding", apart from vague notions of pausing at rare
>codons.

But can be assume that the thermodynamical stability is the same during
translation?  At that time, the chain is terminated on both ends by a
ribosome, the same ribosome.  And the side of the ribosome might
provide a stabilizing substrate.


>>(1) fewer and shorter folding pathways:
>>
>>Interdomain translational pause indicates that in vivo folding
>>happens much faster, in general, than in vitro folding, at close to
>>ribosome speed, in time approximately proportional to chain length.
>>My guess is that it is faster because the in vivo factors eliminate
>>all folding pathways except the one[s] leading to the natural fold.
>>These in vivo factors would make a simulator faster for the same
>>reason.
>
>We have discovered small proteins that can fold in microseconds, and
>these are just on the borderline of being analysed by current MD
>simulation.  The time scale of in vivo folding is 6 orders of
>magnitude slower.

This slowness of in vivo folding surprised me very much, but some
Internet research proved you correct.  Part of the reason seems to be
that the ribosome is *very* slow, operating at 4 to 22 amino acids per
second according to one paper.

Whether intradomain folding needs all the time that it gets is an open
question.  But interdomain translational pause indicates that
interdomain folding needs all the extra time that it gets, and overall
folding time equals translation time plus maybe a little more for the
final domain.

But in any case, you have convinced me that Molecular Dynamic
simulations are unsuitable now for folding large proteins, whether in
vivo or in vitro.  To fold these it might be necessary to resurrect
some of the less compute intensive Energy Minimization methods.  These
might actually work now if the energy function also models the effect
of the attached ribosome.  Or they might not work.


>To summarise what I am saying is that in principle we do need to
>simulate how the ribosome works in order to understand how larger
>proteins fold in vivo, which may lead to a better understanding of in
>vitro process and could speed up in vitro simulation.  And your ideas
>may well be useful for these complex systems. But, we are years away
>in computing power.
>Alan

I guess that my problem is that I am not convinced that we would need
any more computing power if we simulated what we know [already] about
the in vivo process.  But I can not prove that.  I am only a software
engineer that enjoys thinking about science.

I know that you have much better things to do than trying to convince
me.  I thank you very much for giving to me some of your valuable time.
 I have learned a lot from you and the Internet research that you
prompted me to do.  And I wish you great future success.

Thank you again.


Dave.

-
...
----------------------------

-
...
- For info on unsubscribing, etc., please ask
- majordomo@world.std.com for 'help'.

    [Click on (Back) button now.]

    Alphabetical Site Index