First Few Seconds of Evolution

[jerrywickey]

At some point in the far distant past, when earth was devoid of all life, somewhere on the planet, the first short assembly of nucleotides became the first RNA replicator. In the seconds and minutes after that event, this randomly assembled sequence replicated itself over and over again, as fast as it could. This was its only purpose in the universe, to replicate itself.

 

The conditions at that exact point in time and at that exact location were very conducive to the random assembly of nucleotides. Conditions were "hot." So this sequence, perhaps for a short period of time or perhaps extended period, not only replicated itself, but because of the conducive conditions responsible for its spontaneous assembly, it also easily acquired other sequences. Some were advantageous and augmented the original or its daughter's ability to replicate. While others disrupted or stopped completely this ability.

 

If replication ceased, the sequence would quickly be reassimilated in this "hot soup." But none the less, in only minutes or hours, there would be thousands or millions or billions of copies of this first successful replicator RNA sequence. It is this period of time, that I want to see from a front and center seat.

 

I am writing software to simulate just that, the first few seconds of evolution. It will puke out a simplified "RNA genome" of every daughter of the first replicator, while simulating the mutations which must have taken place.

 

I need to double check my parameters. There are program parameters for:

 

base replication rate

frame shift mutation rate

large frame shift mutation rate (i.e. inclusion of a large chunks of RNA)

single nt exchange mutation rate

rate of advantageous RNA sequences

rate of dis-advantageous RNA sequence

rate of fatal RNA sequence

 

We don't need to know the correct values for these variable. The values of these variable can be manipulated to find the set of values which must have represented the first primordial pond, "the fountain of life," But I need some input. I need people's ideas and comments on my model's consistency with observations. I need to know if these are the right variables. Of if I need to replace some with others.

 

Please read Software model of early evolution that really works --

 

http://www.scienceforums.net/forum/showthread.php?t=31212

 

I describe the model in detail there. Post your thoughts. Even an idea that seems nonsense might shake loose another idea which will really help. After all variable mutation and natural selection of ideas works too.

 

Help me out.

 

When I am finished with the software I will post an executable, so everyone can enjoy the show. And I will post the source code, so anyone can play. I am soliciting help with this in this forum and on several other fora. Even fora compete for success.

 

Jerry

[DrDNA]

My first thought is you are making a LOT of assumptions that may or may not be valid.

 

For one, was RNA even first self replicating molecule?

And was replication even the first meaningful act of that molecule? I mean was the first RNA molecule a self replicating ribozyme or did it do other things like make something else first that facilitated the act of replication.

 

Then there is the time issue and the possible role of intermediates. It looks like you are assuming that it was almost instantaneous.

I doubt that it just went: no RNA -> bam RNA -> bam replication....I guess it could have, but I doubt it.

 

Are your "parameters" for ribozymes?

Also, "parameters" would depend on the nucleotide/nucleobase/sugar structure, so I assume you are assuming that RNA structure and composition was then as it is now.

How many base options? four?

I see no "paramater" for environment....eg pH, temp, salinity, buffering capacity of the local environment, amount of oxygen, amount of sulphur, other important ions/compounds/elements, was it done partially on a solid matrix (clay?) or entirely in solution......etc...etc All of your "parameters" depend on these and more.

 

When I think of parameters, I think of independent input variables. But your "parameters" are actually dependant output variables. Maybe it is the difference in chemistry/biology and programming....?

 

It looks like you are trying to model ribozyme activity; one activity of which can be replication....which is a handful all alone. Have you thought about getting that down pat before trying to make the broad leap to biogenisis?

 

EDIT: Oh. I forgot a biggie. With ribozymes, it's all about sequence.

 

It will be arbitrarily assigned as GACUCUUCUCAGGGCC The sequence which allows an organism to reproduce, in my model, is CUCUUCUC. If that sequence is present in an organism's genome, that organism is very very likely to replicate its entire genome, in my world model.

 

1. that is not arbitrary

2. Something this small will self replicate???. It can't bend back on itself. Are you mixing ribozymes with protein mediated replication (DNA and/or RNA reverse trans?)?

Reference please.

Why are you using this, how did it come about, and how are you modeling the higher order structures of this sequence???

[iNow]

Whoa... Nice post, Doc.

[jerrywickey]

DrDNA

 

I am going to address your objections the most important first.

 

1) "Replication is only one activity of ribozyme." While this is true, it is immaterial. Any RNA sequence which does not replicate, will never be plentiful enough to have any lasting effect. It could be a wonderful sequence that does everything including the dishes. But if it can't replicate or be replicated by association with a sequence which can replicate, it is ultimately doomed.

 

My model assumes this replicating sequence as its starting point.

 

2) We already know that the sequence my model uses would do nothing. In fact using the designations U C A G are simply familiar decorations. The assumption is that the model arbitrarily assigns meaning to some sequences.

 

We do not know what the sequence was. We cant even find a self replicating sequence. But if life started with out some sort of extra natural "stirring of the pot," then there must have been one. My mathematical model simply arbitrarily assigns a sequence as the one which accomplishes this.

 

It then arbitrarily assigns other sequences with functionality which augments or reduces the effectiveness of the replication sequence.

 

I do not intend to model chemical activity at all. That would be impossible on that scale with today's desktop computing power. The effects of the mutations will be judged not by a model of their chemical activity but by the arbitrary assignment of function to sequences.

 

I covered this in the post. I hope that clears up your objections to the arbitrary designation and to the possible chemical environment. Simply put, if the chemical environment made it impossible for RNA to replicate, we are suggesting extra natural involvement to facilitate the activity. My model assumes condusive chemistry with out specifying it.

 

3) I will be using four nucleotides. But this is only decorative. It is a mathematical model and the use of A G U C is just a familiarity.

 

4) RNA may not have been the first replicator. Amino adenosine triacid ester was a candidate which MIT proposed. Ultimately it proved unsuitable. But since the world we see today is exclusively DNA, what ever the first replicator was, it must have made a transition to DNA and the only doorway is RNA. Unless we theorize some sort of extreme complex step directly to DNA.

 

This software models that point at which RNA would catalyze its own replication.

 

To quote MIT regarding a the first instance of an RNA replicator. "If emergence of the first RNA replicase ribozyme coincided with the origin of life, it would have had to arise in a single step from prebiotically synthesized RNA, without the benefit of Darwinian evolution. "

 

MIT suggests that if it was RNA, the nucleotides would have had to be assembled randomly. The more difficult questions is getting nucleotides to assemble. There is research which suggests that nucleotides can spontaneously assemble in the presence of proflavin.

 

An organic environment, containing a compound which could act as a "midwife" facilitating prebiotic nucleotide assembly would yield many RNA sequence permutations.

 

We have to start somewhere, unless you are suggesting some sort of extra natural "stirring of the pot." Any replicator must have become DNA based at some point to provide for the current total domination of DNA. There is not another more plausible pathway to DNA than enzymatic RNA.

 

5) "dependant output variables. Maybe it is the difference in chemistry/biology and programming" this is simply semantics. The software doesn't really care what it is called. It will produce the output all the same.

 

 

Do you have any suggestions as to unsuitability of:

 

Assume a replicating RNA sequence whose environment is particularly conducive to random mutation which acquires other sequences which alter its functionality, sometimes advantageously, sometimes disadvantageously.

 

Is this an accurate model in your opinion? And would the results of this model lend helpful insight into the possible patterns of earliest life?

 

 

Jerry

[iNow]

To quote MIT regarding a the first instance of an RNA replicator.

 

<...>

 

MIT suggests that...

 

 

Hey Jerry,

 

How about a specific link or a specific name?

 

 

 

According to Chevy, Fords suck.

[jerrywickey]

The MIT RNA quote:

 

"If emergence of the first RNA replicase ribozyme coincided with the origin of life, it would have had to arise in a single step from prebiotically synthesized RNA, ..."

 

--- Dr.s Wendy K. Johnston, Peter J. Unrau, Michael S. Lawrence, Margaret E. Glasner, David P. Bartel at Whitehead Institute for Biomedical Research, and Department of Biology, Massachusetts Institute of Technology, Cambridge, MA

 

18 May 2001: Vol. 292. no. 5520, pp. 1319 - 1325 DOI: 10.1126/science.1060786

 

 

And the amino adenosine triacid ester (AATE). Take a look at

 

http://w3.mit.edu/newsoffice/tt/1990/may09/23124.html

 

 

This ain't my first Rodeo. I am not trying to pull any punches. I know what I am doing.

 

Would you like to help?

 

Would reviewing a simulation of the first generations of an RNA first replicator be interesting and revealing to you?

 

Jerry

[DrDNA]

I am going to address your objections the most important first.

 

1) "Replication is only one activity of ribozyme." While this is true, it is immaterial. Any RNA sequence which does not replicate, will never be plentiful enough to have any lasting effect. It could be a wonderful sequence that does everything including the dishes. But if it can't replicate or be replicated by association with a sequence which can replicate, it is ultimately doomed.

 

My model assumes this replicating sequence as its starting point.

1. They are not objections. You asked for input and I gave it.

2. Other molecules are not immaterial. In particular if those molecules spontaneously assembled and were an aid to the sythesis of the conformations and sequences that replicated.

 

Often making it too simple, just makes it too simple.

 

2) We already know that the sequence my model uses would do nothing. In fact using the designations U C A G are simply familiar decorations. The assumption is that the model arbitrarily assigns meaning to some sequences.

Then I am missing your point. And what ever you get out is only going to be as good as what you put in.

We do not know what the sequence was. We cant even find a self replicating sequence. But if life started with out some sort of extra natural "stirring of the pot," then there must have been one. My mathematical model simply arbitrarily assigns a sequence as the one which accomplishes this.

 

It then arbitrarily assigns other sequences with functionality which augments or reduces the effectiveness of the replication sequence.

I do not intend to model chemical activity at all. That would be impossible on that scale with today's desktop computing power. The effects of the mutations will be judged not by a model of their chemical activity but by the arbitrary assignment of function to sequences.

I must not get it because it looks like, without taking into account any of the chemistry, you are trying to determine function...which is a chemical function.

Replication is a chemical activity.

 

I covered this in the post. I hope that clears up your objections to the arbitrary designation and to the possible chemical environment. Simply put, if the chemical environment made it impossible for RNA to replicate, we are suggesting extra natural involvement to facilitate the activity. My model assumes condusive chemistry with out specifying it.

What "extra natural involvement" are you refering to?

I am more interested in what chemical enviroments would have made it possible.

For one thing it is difficult to prove a negative.

3) I will be using four nucleotides. But this is only decorative. It is a mathematical model and the use of A G U C is just a familiarity.

 

4) RNA may not have been the first replicator. Amino adenosine triacid ester was a candidate which MIT proposed. Ultimately it proved unsuitable. But since the world we see today is exclusively DNA, what ever the first replicator was, it must have made a transition to DNA and the only doorway is RNA. Unless we theorize some sort of extreme complex step directly to DNA.

Whose is MIT?

RNA may be a doorway now, but you don't know what doorways existed when the atmosphere and the environment were nothing like the oxygen rich, near neutral pH atmosphere we enjoy today.

 

This software models that point at which RNA would catalyze its own replication.

But unless you have developed an omniscent intelligent being, that software will only output information that is based on what you put in.

I am trying to determine exactly what it is you are putting in.

 

To quote MIT regarding a the first instance of an RNA replicator. "If emergence of the first RNA replicase ribozyme coincided with the origin of life, it would have had to arise in a single step from prebiotically synthesized RNA, without the benefit of Darwinian evolution. "

 

MIT suggests that if it was RNA, the nucleotides would have had to be assembled randomly. The more difficult questions is getting nucleotides to assemble. There is research which suggests that nucleotides can spontaneously assemble in the presence of proflavin.

 

An organic environment, containing a compound which could act as a "midwife" facilitating prebiotic nucleotide assembly would yield many RNA sequence permutations.

 

We have to start somewhere, unless you are suggesting some sort of extra natural "stirring of the pot." Any replicator must have become DNA based at some point to provide for the current total domination of DNA. There is not another more plausible pathway to DNA than enzymatic RNA.

You are going to give a specific reference. There are too many profs at MIT to count.

 

5) "dependant output variables. Maybe it is the difference in chemistry/biology and programming" this is simply semantics. The software doesn't really care what it is called. It will produce the output all the same.

You may consider it a minor detail. I do not. For one thing, when you confuse dependant and independant variables, the integrity of the data is lost.

 

Do you have any suggestions as to unsuitability of:

 

Assume a replicating RNA sequence whose environment is particularly conducive to random mutation which acquires other sequences which alter its functionality, sometimes advantageously, sometimes disadvantageously.

 

Is this an accurate model in your opinion? And would the results of this model lend helpful insight into the possible patterns of earliest life?

 

 

Jerry

It sounds obvious. Except it appears you are trying to come to conclusions about biology and chemistry by ignoring the laws of chemistry and physics.

 

What is your objective?

 

This ain't my first Rodeo. I am not trying to pull any punches. I know what I am doing.

 

Would you like to help?

 

Would reviewing a simulation of the first generations of an RNA first replicator be interesting and revealing to you?

 

Jerry

 

Thank you but no. At least not yet anyway.

 

I want to know what is going in before I see what is coming out.

 

I'm really curious as to what it is you wish to accomplish with this.

[Realitycheck]

At first I thought he was referring to an article I pointed out, but then I figured that surely they are way ahead of this currently and found some of the stuff they are publishing these days. Here is a search of MIT using "abiogenesis". Of course, all of this stuff is pay-per-view.

 

http://www.mitpressjournals.org/toc/artl/14/1

 

 

The Renaissance of Synthetic Biology

Juli Peretó, Jesús Català

Biological Theory Spring 2007, Vol. 2, No. 2: 128-130.

 

Genetic Exchange Leading to Self-Assembling RNA Species upon Encapsulation in Artificial Protocells

Sergio-Francis M. Zenisek, Eric J. Hayden, Niles Lehman

Artificial Life Summer 2007, Vol. 13, No. 3: 279-289.

 

 

Then, I did a search on Rebek, who did the study that I pointed out, and came up with these other articles.

 

 

Fifty Years of Research on Self-Replication: An Overview

Moshe Sipper

Artificial Life Summer 1998, Vol. 4, No. 3: 237-257.

 

Evolving Reaction-Diffusion Ecosystems with Self-Assembling Structures in Thin Films

Jens Breyer, Jörg Ackermann, John McCaskill

Artificial Life Winter 1998, Vol. 4, No. 1: 25-40.

 

Evolvable Self-Replicating Molecules in an Artificial Chemistry

Tim J. Hutton

Artificial Life Fall 2002, Vol. 8, No. 4: 341-356.

 

So I guess this is besides the point of the limited scope of the software, but if we get this perfect little strand of rna/dna that has it all mapped out about what to do and winds its way up in a little cell-like structure, then it can go to work and, behold, we have a first cell. The odds of this happening make the chances of life occurring extremely, infinitely small, unless there is an easier way for it to be accomplished.

 

How else could it happen other than by chance, serendipity?

 

Here are a few other noteworthy articles.

 

Synchronization Phenomena in Surface-Reaction Models of Protocells

Roberto Serra, Timoteo Carletti, Irene Poli

Artificial Life Spring 2007, Vol. 13, No. 2: 123-138.

 

Live Evolving: Molecules, Mind, and Meaning. Christian De Duve. (2003, Oxford University Press.) Hardback, £25, $39. 358 pages

Carlos Gershenson

Artificial Life Winter 2007, Vol. 13, No. 1: 91-92.

 

New Wine in Old Bottles

Evolution: From Molecules to Ecosystems

Andrés Moya and Enrique Font, eds

Oxford: Oxford University Press, 2004 (350 pp; $185.00 hbk; ISBN 978019851425)

 

An Approach to Biological Computation: Unicellular Core-Memory Creatures Evolved Using Genetic Algorithms

Hideaki Suzuki

Artificial Life Fall 1999, Vol. 5, No. 4: 367-386.

 

Chemical Genetic Algorithms—Evolutionary Optimization of Binary-to-Real-Value Translation in Genetic Algorithms

Hideaki Suzuki, Hidefumi Sawai, Wojciech Piaseczny

Artificial Life Winter 2006, Vol. 12, No. 1: 89-115.

 

 

Life Cycle of a Minimal Protocell—A Dissipative Particle Dynamics Study

Harold Fellermann, Steen Rasmussen, Hans-Joachim Ziock, Ricard V. Solé

Artificial Life Fall 2007, Vol. 13, No. 4: 319-345.

[DrDNA]

The MIT RNA quote:

 

"If emergence of the first RNA replicase ribozyme coincided with the origin of life, it would have had to arise in a single step from prebiotically synthesized RNA, ..."

 

--- Dr.s Wendy K. Johnston, Peter J. Unrau, Michael S. Lawrence, Margaret E. Glasner, David P. Bartel at Whitehead Institute for Biomedical Research, and Department of Biology, Massachusetts Institute of Technology, Cambridge, MA

 

18 May 2001: Vol. 292. no. 5520, pp. 1319 - 1325 DOI: 10.1126/science.1060786

 

 

And the amino adenosine triacid ester (AATE). Take a look at

 

http://w3.mit.edu/newsoffice/tt/1990/may09/23124.html

 

 

Jerry

 

That is a Chemistry (bioorganic chemistry) paper Jerry. Rebek is very well respected in this field. So how are you not taking into account chemistry?

[jerrywickey]

Modeling the chemistry would require Sci Fi like computing power. Modeling with out addressing the chemistry allows the software to replace chemistry with logical math. Similar to popular complex war games where two army's fight based on the role of a die by weighting the die's outcome with considerations for the opposing army's strengths and weaknesses.

 

The arbitrary sequence I chose to endow reproductive characteristics is ridiculously small. This is obvious, reading the MIT paper on the 195nt replicase. But the software does not model chemical activity. It models the theorized effects of small mutations on sequences which ultimately endow relatively complex functionality.

 

In the resulting runs, I expect to see patterns that are similar to what millions of years later will become protein evolution.

 

I expect to see patterns of development which show overlapping function of sequences. This is one of the most exciting processes I hope to observe.

 

Else where in these posts I comment on other things I hope to the model will illuminate.

 

Jerry

[jerrywickey]

Dr DNA

 

Sorry I did not have time to respond more carefully before.

 

You might be missing my point. I don't know. I am attempting to see the emergent patterns of early replications. I am not trying to solve the chemistry.

 

One of the most exciting things I found the very first night I completed the program and the most interesting as well to Dr. Aller was that it turns out that replication alone in an environment conducive on.y to mutation with no possibility of advantageous or disadvantageous activity will by itself generate genome growth.

 

This revelation could only have come from a model where the parameters can be completely controlled.

 

Neither I nor Dr. Aller could understand the mechanism. The output of the software made this phenomena obvious. But neither of us realized it would happen.

 

What comes out is not what I put in. What I put in are only the rules which govern the process. The process unfolds according to those rules and often surprises us.

 

Being able to model the chemistry directly would be wonderful. If we had the computing power to do this, we would already know exactly what the first replicator sequence is. We would already know exactly what the conducive chemistry was.

 

But we don't so we still do not know.

 

I am not referring to any extra natural involvement. I am suggesting that if e.n.i. was not involved then an accurate model must be available.

 

MIT is Massachusetts Institute of Technology. And the chemistry then was the same as chemistry is now. We will not likely find that some sort of new chemistry was operating long ago. DNA is what the earliest replicators must have gravitated toward, because DNA is all we see now.

 

We can sufficiently hypothesis about what conditions might have been long ago. And we can investigate those conditions.

 

However, any replicator of any kind under any conditions, must have made a switch to DNA at some time. Unless we presume an implausibly complex molecular system that jumps to DNA, RNA is our only option.

 

Do you have another in mind or are you supposing that Darwin's black box holds another secret possibility?

 

The motivation for writing this software was in fact my review of other evolution software. All I reviewed is exactly as you describe. Nothing comes out that they did not put in.

 

Dr. Aller and I spend several hours testing my software against his lab observations. My software accurately duplicated every empirical lab result, we tested.

 

I have confidence that it will predict accurately otherwise untested values for parameters.

 

I used simple logical constructions, the output is an emergence of these simple rules. The results of true emergence often surprise the author of the simple rules. This was exactly the case when the Dr. and I observed the sim engines telling us that replication alone with out advantageous mutation did in fact increase genome size.

 

When we tested advantageous mutations only, it grew faster. And grew fastest with advantageous mutations. We expected this but the growth with out advantageous or disadvantageous activity was not input and was a surprise.

 

It took us an hour, to discover the bio mechanism which accounts for this. This the first thing this model taught me..

 

Jerry