|These emails, mostly from Brig Klyce to Tom Ray, are a few of the 500+ exchanged during a research project at the University of Oklahoma, funded by the Astrobiology Research Trust. At the project's inception, Mark Bedau of Reed College was the point person at OU. Freemen Dyson's opinion was solicited once, and Evan Eichler later collaborated on the project. A paper on it was presented at NASA's Astrobiology Science Conference 2002, April 7-11. These emails, in chronological order, reveal the development of the research project.|
Correspondence with Tom Ray and others, 2001-2005
Date: Wed, 22 Aug 2001 17:24:50 -0500
Dear Mark --
Yesterday I spoke with Tom Ray about the Astrobiology Research Trust. He suggested that I get in touch with you also. He said that you and he plan to collaborate on issues that relate to my interest. I am establishing a non-profit agency to fund research pertaining to astrobiology -- hence the name. In the first year or so, I plan to fund the agency myself. Later I would like for it to grow and acquire a broader base of support.
While we are still small, we will have the best chance to make a difference if we focus on few or even only one issue. I am personally interested in an issue that broadly overlaps with Tom's work with computer models like Tierra.
Allow me to paraphrase Tom's question as "In artificial life, how is evolutionary progress in a closed system possible?" He would like to be able to model the evolutionary progress exhibited by real life on Earth. In my opinion, the advent of astrobiology changes the situation for real life on Earth. If the planet is not a closed biological system, the previous evidence is likely contaminated. This raises a radical possibility. What appears to be evolutionary progress in real life on Earth may actually be the local _development_ of previously existing highly evolved life from beyond Earth.
Please hang on!
And so I would like to consider the question, "In real or artificial life, is evolutionary progress in a closed system possible?"
I am aware of Richard Lenski's closed system biology experiments, and the results so far, in my opinion, do not provide a positive answer. Because one problem could be that those experiments simply haven't run long enough, computer models look attractive to me.
May I assure you that I have no interest in creationism. I think some creationists raise valid points against the Darwinian account of evolutionary progress; but the alternative I promote, strong panspermia, is entirely scientific, or materialistic, if that is clearer. I do run afoul of big bang zealots, because I maintain that the big bang theory isn't secure enough to close this subject.
And I promote strong panspermia as an alternative to Darwinism without being sure that it is correct. Rather, I have come to accept the thesis that a theory can only be replaced by another theory, and I do not want to completely lack an alternative.
However, when Michelson looked for evidence of the luminiferous aether, he was simply trying to confirm what everyone already believed, just double-checking, so to speak -- not promoting any other theory. A major scientific advance followed soon after the definitive Michelson-Morely experiment.
I suggest that my question is similar, just double-checking. If evolutionary progress in a closed system is possible, one ought to be able to demonstrate it in real-time experiments.
If the answer is yes, the existing paradigm would get a much-needed boost. If the answer is no, we don't have to all become creationists, as both Darwinists and creationists wrongly assert.
I am aware that there are issues that may have to be resolved before one can tackle the question head-on, such as, how do you recognize or measure evolutionary progress? Such issues are worthy on their own of course.
Some of my webpages about strong panspermia make some of the same points as above. If you like, see http://www.panspermia.org/computr2.htm and the two "Next" pages after it.
I am just a guy who wants to find out what's going on, because I think it matters. This question and the Astrobiology Research Trust are the best way I can think of to participate and to try to help find out.
If you and Tom are interested in any of this, I would like to discuss with you the possibility of the Trust's supporting you and / or the new center that Tom mentioned when we talked. Will you please share this email with Tom?
I will hope to hear back from him or you after you have considered this. Thank you for your kind attention. Best Regards.
Date: Fri, 5 Oct 2001 10:31:43 -0400 (EDT)
Dear Brig Klyce,
Thank you for today's message. Now I see clearly what you have in mind. Unfortunately I do not think your suggestion makes sense. We know several examples of situations where we can see evolution happening, and many more examples where we can see clearly the results of evolution happening in the recent past. One example is Tom Ray's community of digital creatures evolving by means of his Tierra program. Another is the famous case of populations of moths turning black to avoid predators in nineteenth-century England when trees were blackened with soot, and then turning pale in the twentieth century when the environment was cleaned up and the soot disappeared. Since we know that there are situations where evolutionary progress happens, it does not make sense to ask whether evolutionary progress is possible. That question is already answered.
Another beautiful example of evolutionary progress is the evolution of whales from the cow-like creature recently discovered in Pakistan and pictured on the cover of this week's Science magazine. The evidence that the Pakistan fossil is ancestor to whales seems to me quite compelling. And the evolution of many species of majestic whales from a cow is undeniable evidence of progress.
I am sorry that I find your program unhelpful and I have no interest in supporting it.
Yours sincerely, Freeman Dyson.
Date: Sun, 21 Oct 2001 23:54:04 -0500
[Tom writes] I could not tell from your response if you consider the emergence of humans from apes to represent progress. However, it appears from your summary that you do not feel that the evolution of humans from apes requires strong panspermia.
Dear Tom -- Thank you for pressing me to be clear. The emergence of humans from apes absolutely represents progress. I am nearly certain that this step requires strong panspermia as a proximate cause. I believe this because I have a strong hunch that humans express genetic programs that apes do not express. The "language gene" might be an example or a part of such a program. The installation of this program into the ape-human lineage, whenever that occurred, would exemplify the working of strong panspermia.
I was equivocal because the installation might occur long before the expression of the genetic program. Probably too equivocal, because strong panspermia would still have to have provided the program, only earlier.
And for all I know about primate genetics, it may be that apes actually express all the genetic programs that humans do. (In this case, the "language gene" would somehow be expressed in apes.) But if brain size accounts for all the differences, for example, and if brain size depends on something like the frequencies of alleles that both species express, then strong panspermia would have no immediate role in this evolutionary step. Darwinism would be sufficient for the final "sorting out." This latter equivocation may seem to open the door for Darwinism to account for all of evolutionary progress, but it does not. Very many steps in evolutionary progress cannot be accounted for by allele frequencies, we already know.
Do geneticists know if humans express or even posses genetic programs that apes don't?
Thank you for your continuing interest. I hope I am making sense to you. If not, please press for clarification. Thanks. Brig
Date: Mon, 22 Oct 2001 17:30:00 -0500
Dear Tom --
Many of them, like body size or proportions, and amount or length of hair, might conceivably arise by Darwinian evolution without new genetic programs.
> I understand your position that the genes that are responsible for
The chance for genes to be installed before they are expressed does not make my theory untestable. It makes it more difficult to test in highly evolved species, because the genomes are large and hard to understand (and the system is open). That's why I earlier contemplated closed tests in simpler systems like bacteria or computer models where the initial genome can be thoroughly inventoried and external sources controlled.
BTW, my original request was to test not my theory, but Darwinism -- Can it produce progress in a closed system? No loopholes there. I am pleased that you have another idea. When you asked me about it, I wanted to be sure to say everything that might be relevant. On this subject I do not lack courage.
> What I propose is to compare the entire human genome to the ape
Transfer from anywhere supports strong panspermia, right? Darwinism is supported if you find incremental pathways -- consisting of plausible mutational steps like duplication and divergence -- leading to progressive new genetic programs from genetic material without those programs.
> From a broader perspective we can view it
> I expect that we will find new, uniquely human gene families.
I do, too. But what would you conclude from that? New, uniquely human gene families emerged how?
> If not, then we will need to look more deeply into the
Yes, or pick another example where new gene families are already known to have emerged.
> We must determine how gene families emerge.
Knowing how new gene families (or new genetic programs) emerge will probe strong panspermia and Darwinism.
> Do you want to support this research?
Yes, I would like to support research that seeks to discover how gene families emerge. I would to the extent that --
Are you comfortable with these clauses?
I have thought some more about the subject of contract research. I thought you don't do it, but you told me you do. Would this project go better if we think of it that way? Well-written contracts can prevent misunderstanding. I could try to draft one, if you think that would be useful. The essence of it would be the two clauses above. In any case, I believe we can make this collaboration work. Thanks again for thinking so carefully about it. I apologize if I'm frustrating. Lets' keep talking.
Best regards. Brig
Date: Fri, 02 Nov 2001 17:59:12 -0600
Dear Tom and Mark --
Thanks for the good visit. I enjoyed it and I thought lots was accomplished. I am waiting to see and sign a memorandum of understanding, as I recall. More than ever I want this cooperation to work.
Meanwhile, Tom, I have a followup question about the genomics investigation. You said that if strong panspermia drives evolution, the signs will be obvious and it will be big news. But remember the paragraph I showed you from Nature (H. Ochman, et al. "Lateral gene transfer and the nature of bacterial innovation," Nature 405, pp. 299-304, 2000) --
"It is difficult to account for the ability of bacteria to exploit new environments by the accumulation of point mutations alone. In fact, none of the phenotypic traits that are typically used to distinguish the enteric bacteria Escherichia coli from its pathogenic sister species Salmonella enterica can be attributed to the point mutational evolution of genes common to both. Instead, there is growing evidence that lateral gene transfer has played an integral role in the evolution of bacterial genomes, and in the diversification and speciation of the enterics and other bacteria."
As I understood your criteria for evidence of strong panspermia, they are already met in bacteria. But you yourself seem not at all impressed. So, I must be missing something.
Enough business. ...Thanks guys. Reply is optional, unless Tom can help me with the bacteria issue. Brig
Date: Thu, 08 Nov 2001 22:49:46 -0600
[...] Tom, I honestly think the likely reaction among mainstream scientists has already been demonstrated by the bacterial situation. No one is clamoring for a new explanation there. And the same situation -- growing evidence of transfer and little for Darwinism (as a means to make progress) -- is already beginning to emerge for eukaryotes. However, if, before your research, the question is posed, and some respected Darwinists go out on a limb and predict that some percentage of new genetic programs in humans are internally generated, then the impact of the result should be enhanced.
[Do you remember the question I asked you earlier -- what if you find uniquely human gene families with no obvious source (besides transfer from an anonymous species)? What will you think or do then?]
I'm liking the prize more and more. I would expect lots of different approaches to be proposed. And maybe a conference would produce some good ideas. Mark mentioned a conference.
I do not feel that it is necessary to nail down a research method before we have a memorandum of understanding. If the Center makes testing the possibility of evolutionary progress in a closed system one of its research priorities, we'll find a way.
Sorry not to have responded immediately. I started one but it annoyed even me!
Date: Fri, 09 Nov 2001 15:30:19 -0600
The first appearances of oxygen metabolism, photosynthesis and flagellae are examples of progress among bacteria. But if bacteria can't make progress in a closed system, are they different from what comes both before and after them? If so, how did anything more advanced manage to come after them?
However, if you want to examine an ancestral species and a descendant species, I can see why bacteria are problematic.
> The current conventional wisdom says that in higher organisms
I agree with you! Will others?
The case could be made that a Darwinian process wrote the new genetic program elsewhere, and then it was easier for humans to get it by transfer than by re-writing it from scratch. A similar argument is already made with respect to the gene that controls eye development across species as different as wasps, squid and mice. See for example
But my primary reaction is this -- if you like your method and if you think it will convince others, great. My first goal is for the question to be taken seriously. You seem to take it seriously.
I am still standing by for a memo -- primarily for your benefit, as I understand it. Is there anything else I can do help expedite this process or the establishment of the Center?
Brig Klyce / Acorn Enterprises LLC
Date: Sat, 24 Nov 2001 10:25:48 -0600
Dear Mark --
Thanks for your last. I am in Montana (where I do not have my previous emails from you.) I am indeed still interested in establishing the prize.
I have two concerns about it. One is defining the criteria so that the awarding of it would definitely mark a scientific advance. The other is keeping it from becoming circus-like. The mere existence of the prize would surprise some people, wouldn't it?
Anyway, yes, I would like to pursue the idea. I would be interested in your preliminary budget also. And I would like to find another word besides "progress." Any suggestions?
Best regards. Brig
Date: Mon, 3 Dec 2001 14:00:20 -0800
Thank you for your thoughts of last Friday. Let me respond to the key issues. The more complex issues are at the end.
Sure, we would announce much more widely than in the Artificial Life journal. Though it would be good to have the journal be the official clearing house.
I agree with you completely. It would be easy to make the distinction clear, and it would be quite natural to explain that we need to answer your question *first* before turning to questions like "How life arises from the non-living". In fact, your question is also directly related to question #6 "What is inevitable in the open-ended evolution of life?"
> > Yes, we need to find another word besides "progress". I would want
> I agree. One example gives me pause however. In Conway's "life" there
One key problem with Conway's "life" is that the complex outcomes it generates are not "robust"--they do not repair themselves, so they are brittle and fall apart at the smallest disruption. In this respect, they are quite unlike life. We could deal with this by adding some word like "robust" to the description. To deal with the genotype/phenotype issue, we could add that the requirement that the creation process involves evolution. This gives us the following wording: "repeated instances of the evolution of robust, qualitatively new and more complex adaptive structures....
You told me on the phone recently that you think that artificial evolution has already had enough of a chance to prove itself. I don't agree. I think that there is still an important untried hypotheses about how artificial evolution in a closed system could create repeated instances of the evolution of robust, qualitatively new and more complex adaptive structures ("progress", for short). The central idea is what I outlined in my email on 29 Nov: having evolution occur in a population of entities that are self-organizing and self-maintaining. (By the way, this make the structures "robust".) I don't know if this will do the trick, but it is premature to conclude that it won't work. It's still an open question.
A further important complication is that objectively measuring the evolutionary potential of this sort of system is difficult, precisely because the evolving structures are reorganizing themselves through the evolutionary process. To dramatize the point, the problem is to observe the creation of new "widgets" when you don't know in advance what kind of "widgets" might emerge. (I call this the emergence problem.)
Work on developing appropriate measures for evolutionary creativity must solve the emergence problem. And it is most fruitful to develop these measures by testing them in systems in which new "widgets" emerge in the evolutionary process. Since these models do not yet exist, I don't see any realistic way to develop, test, and refine the appropriate measures without simultaneously creating new artificial systems. This is way trust (i) developing measures of evolutionary "progress" is essentially connected with thrust (ii) developing new artificial evolutionary systems. I don't see how to do thrust (i) without also doing thrust (ii).This explains why I have always combined thrust (i) with thrust (ii), ...(In truth, even refining the measures with *existing* systems would involve lots of simulations, which involves graduate assistance and simulation hardware and software.) This also explains why measuring evolutionary "progress" as needed for the prize is a non-trivial job. Even creating a appropriately defined prize would be a significant scientific advance, as I mentioned earlier. It is not reasonable to expect this to be done in time to announce at the upcoming astrobiology conference in April, nor AAAS in February, nor the next GECCO. It sounds like you want to create the prize very quickly. I don't think this can be done.
When we first met, I remember telling you that I was impressed that you were trying to push the strong panspermia angle by supporting efforts to *falsify* it. Many people want to pursue their individual interests by seeking only evidence that supports their preconceptions. You were willing to also seek for evidence that would undermine strong panspermia--in the hope of most quickly finding out the truth however it comes out. In the arena of artificial evolution, I think this is still the best way to shed light on strong panspermia (in addition to calling people's attention to the question) is still to pursue the most promising efforts to create evolutionary progress in artificial evolutionary systems. The best way I can further strong panspermia is by trying seeing whether I can make an artificial evolutionary system show real evolutionary progress. But perhaps you no longer want to go down this path. If you don't, I'll understand, but I'm not sure how I can be of much help to you.
When you've thought this over, let me know your reaction. Feel free to call me on the phone.
Date: Wed, 05 Dec 2001 15:51:47 -0600
Dear Mark --
I did not mean that people should stop trying. What I mean is that it is time, now, to alert the public, and biologists who care about evolution, that artificial evolution has not yet succeeded in confirming [sustainable evolutionary progress in a closed system]; and that it may or may not ever succeed. (I was extremely gratified when both you and Tom seemed to agree with me about this.) The public and most biologists are currently misinformed on this issue. That is a big reason, perhaps my primary one, for establishing the prize.
So anyway, I would prefer not to wait for one last thing to be tried first. Of course I understand if that is your preference. But you say,
What if, instead of "helping them figure out how to define the prize in objective terms," you said, "helping them *begin to* figure out..."?
> A further important complication is that objectively measuring the
See, I can imagine announcing the prize right now. The criteria at this point could be that the model should "clearly exhibit the capability to make evolutionary progress in a closed system, in the manner that life on Earth appears to have (if one assumes Earth is a biologically closed system.)" The announcement could also say that the criteria will continue to be made more precise by the panel and by discussion with other experts including contestants. But a model that would win under the above general criteria is easy to imagine, and could emerge to claim the prize immediately. I give an example in my poster at NASA's April 2000 Astrobiology conference.
BTW, I have misgivings about the word "emergence." In Conway's Life, gliders and other persistent forms are examples of emergent properties, right? But these are merely surprising behaviors that have nothing to do with the problem. If new widgets arise from new instructions, these are more than emergent properties, I think.
Furthermore, with respect to widgets, they need to help their organisms survive, right? In new, unanticipated ways, right? In fact, I think you can't know in advance what the widgets might be. Defining the widgets as those your system found wouldn't be fair to other contestants.
But more generally, here's where I am --
But when Tom agreed that a predominance of horizontal transfer would constitute evidence for strong panspermia, I was encouraged. The biology question is my main interest, after all. Now I have made a 2-year commitment to his project which is, annually, more than I had planned. I still want to support the prize.
I think I agree that the first problem is how to establish the criteria. But I think that should be done by a panel, somehow representative of all possible approaches, that administers the prize....
Thanks. best Regards. Brig
Date: Sat, 08 Dec 2001 11:06:12 -0600
Dear Tom --
NASA's 2nd Astrobiology Conference is 7-11 April, 2002, at NASA Ames. I went to the first one, 2 years earlier, and it seemed like a success. Some people who might interest you will probably be there. Last time, Peter Gogarten, an expert on horizontal gene transfer was there. And Andrew Pohorille, head of NASA's Center for Computational Astrobiology was there. (But the NCCA may have died. I can't find it on the Internet.)
Do you have any interest in writing or co-writing a paper or poster for this one that explains our new project? It could be an elaboration of the research plan you wrote. Or merely that thing as is. If you don't want to go, I could go and man the poster. I would enjoy that. There would surely be some interesting contacts there. What do you think? NASA seems slow in getting out the details of the conference, but there is a webpage at http://astrobiology.arc.nasa.gov/conferences/2001/ABSciConf/
Date: Sun, 09 Dec 2001 17:59:25 -0600
Dear Tom -- After being sick at home all weekend, today I finally felt good, and I wrote a rough draft for a paper. The purpose of the paper would be to explain our project, and generate interest, comments, participation or whatever. The paper could be presented at NASA's conference in April that I mentioned earlier.
It's almost just a concept at this point, but it's far enough along for you to have a look and react to the concept. Most parts need lots of fleshing out. A really speculative part could be added extending the graphs back to the assumed time of the big bang. Anyway, you can look at
http://www.panspermia.org/graphspaper.htm [now the final version]
I will also invite Dan McShea to look and comment. He and I looked at similar graphs a couple of years ago with fruitful discussions. You won't hurt my feelings ... -- Brig Klyce
Date: Wed, 09 Jan 2002 23:10:18 -0600
Dear Tom -- Will you please have a look at the 4 paragraphs below. These are my proposed [replacement] opening paragraphs of our paper. They give it a softer intro geared to the astrobiology audience. In the later paragraphs, I think how much to say about the specifics of the plan is up to you. Of course, the original purpose of the writing was to establish an agreement between the two of us after much discussion. The audience for this paper is different.
My goals are to alert the astrobiology crowd about our plan, put our expectations on record, and possibly generate interest from another researcher or two -- maybe an A-lifer or bacterial genomics person like Lenski, for example.
Please comment. Thanks.
Date: Tue, 29 Jan 2002 11:05:36 -0600
Dear Tom --
Assuming NASA accepts our abstract, I would like to present a poster that embellishes the story told in the abstract. I sent you, some weeks ago, a very rough draft of a paper to present,
http://www.panspermia.org/graphspaper.htm [now the final version]
but you said you disagreed with it or parts of it. Before I start over, it would help me to know what you disagreed with about it. Can you tell me in the manner that is most convenient for you? Email, telephone, or something else. At the same time, is it too soon to discuss what evidence from the genomic analysis you think will support Darwinism or strong panspermia? I would like to try to be able to discuss this intelligently with you before the results start coming in.
Date: Wed, 13 Feb 2002 22:42:50 -0600
Dear Tom --[…] I'm glad you think the poster is OK. My Memphis advisor, Stan Franklin, asked some "dumb" questions about it, and I would like to try to answer them. This will make it slightly longer, which it needs anyway. I'll show you any proposed changes for approval. The conference is not until early April, so there's no rush. Don't worry about me while you're doing the Japanese thing.
I am thrilled that the work of identifying uniquely human genes that emerged through duplication and divergence is nearly complete. Holy smoke. If you make a preliminary report, I'd like to be there.
Remember Damon Lisch, who wanted to contact you? I have just re-read a 1997 article by him and Kidwell, about transposons (for a colloquium honoring Dobzhansky, I believe.) I think they describe, without meaning to, a way for new genetic programs to become installed as cosmic ancestry would require. I'll send it to you in case you have interest and time for it.
Best regards.... Brig Margaret G. Kidwell and Damon Lisch, "Transposable elements as sources of variation in animals and plants" [text], p 7704-7711 v 94, Proc. Nat. Acad. Sci., USA., July 1997.
Margaret G. Kidwell and Damon Lisch, "Transposable elements as sources of variation in animals and plants" [text], p 7704-7711 v 94, Proc. Nat. Acad. Sci., USA., July 1997.
Date: Tue, 09 Apr 2002 22:34:30 -0500
Dear Tom --
At Moffet Field today, Chris Adami presented his work, mainly "Survival of the Flattest," after which I introduced myself and showed him our poster. I also told him about the challenge I want to pose to the AL community -- to demonstrate evolutionary progress in a closed system. He said that he had already done it in a [1999?] paper for PNAS, i.e. defined complexity [for the first time] and shown why its increase is inevitable. After some discussion of the definition of complexity, I told him that you said it hadn't been done, which phased him not at all. But he acknowledged a debt to you for pioneering work in the field.
Later I drove to and beyond Los Gatos to meet Bruce Damer and his partner Galen Brandt -- a very companionable female vocalist, Yale '75. I toured their residential compound and vintage computer museum, which includes a Cray I. Bruce was far more laid back or unegotistical than I expected after our emails and phone conversations. We talked for a couple of hours. I liked him and, in any case, I am desperate for allies. I said I would email him from Montana in a few days.
Poster traffic was very light tonight, not just for us. Hangar One, where the meeting is housed, would hold, very snugly, three Titanics side-by-side, with room left over at the ends. Birds residing there make much racket in the mornings, although I haven't seen them.
Best regards. Brig Klyce
Date: Tue, 09 Apr 2002 22:34:44 -0500
Dear Tom --
...As a first pass, I think the Darwinian prediction is as follows --
After "uniquely human genes" (versus the mouse) are identified, increasingly primitive versions of them will be found, as one goes farther back in time to the intervening genomes (ape, chimp, lemur, etc.) They will appear to have been constructed bit-wise (base-wise), in the manner of "METHINKS IT IS LIKE A WEASEL," in Dawkins's _The Blind Watchmaker_.
The strong panspermia prediction is that the genes, if found at all, will be nearly the same as finally, with only a few point mutations and possibly some interruptions into a few large pieces versus the human version. Some of the large pieces might be missing, even from the latest predecessor. (The pieces could be reassembled in humans by the genetic version of syntax- and spellcheck.)
Incidentally, might the "uniquely human" portions be less or more than whole genes, like single exons or distributed groups of exons?...
Date: Fri, 12 Apr 2002 10:57:32 -0500
Tom, thanks for the PNAS reference. Also, thanks for telling Lenski about our project. I'm surprised he remembers my website, because he was standoffish and suspicious in my email correspondence with him of almost 2 years ago. Anyway, you are obviously a much better ambassador for our collaboration than I am.
I spoke with W. Ford Doolittle briefly at the NASA conference. He cautioned me that most of his work on transfer pertains to prokaryotes. He said that he expects very few human/nonmouse genes to come from transfer. I told him that I have quoted his "many eukaryotic genes... seem to have come from nowhere," and he said, "But I didn't really mean it!"
I will call you soon. Brig Klyce
Date: Sun, 14 Apr 2002 19:40:12 -0500
Dear Tom --
At the NASA conference there were complimentary copies of 'OMICS: A Journal of Integrative Biology'. An article by Julius H. Jackson et al. (pp 115-121, v6 n1 (2002)) suggested something interesting -- although sequenced bacteria always have a large number of ORFs that can't be recognized, these guys at Michigan State think most of them are functionless.
If the search for matches doesn't already include eukaryotic genomes, I wonder what matches might turn up if it did. I have asked Jackson this question by email just now....
Thanks. Brig Klyce
Date: Wed, 30 Oct 2002 17:14:43 -0600
Dear Tom --
Firstly, many thanks to you -- and Chen Mei -- for hosting me so royally there in Norman. It was an especially great pleasure to enjoy a beer and fresh fruit (my two favorite foods) in your home, while listening to learned guys like you and Evan and your associates. Thanks also for your recent emails. I was in China for two weeks and then catching up for 3-4 days until now. I have some comments --
> I thought it was interesting that Long confirmed, with
The phenomenon you describe does not contradict my theory. In fact, mutation is a necessary part of the theory, for exploring within a narrow range (such as optimising color sensitivity in an opsin).
The way to render my theory superfluous is to firmly establish, in closed system experiments, the ability of life (or a system analogous to it) to make sustainable evolutionary progress (not sideways or backwards, which directions are not in dispute.) Also good would be to convincingly show, step-by-step, how some genes arose de novo from sequences completely unlike themselves, as you were trying to do.
There are lots of reasons to accept the existing theory, just as there were lots of reasons, before Michelson-Morley, to believe in a detectable aether. But until the thing is observed, there's room for doubt. It would really delight me if I could convince you that the lack of confirmation [of evolutionary progress] in closed systems is a serious problem.
I sent an email to Long and copied you. I wish I coulda been there in Norman when he came. Please invite me to the next ones. Thanks also for the two article references. One, "Genes caught skipping from bacteria to beetle," I had seen announced somewhere. I have accumulated many dozens of examples of genes hopping across species, kingdom, domain etc., boundaries, so I do not doubt it at all. The other one reports, "Now two research teams in the United States have produced evidence to rebut the claim [horizontal transfer from bacteria to people.]" This may become the object of a lengthy debate, but genes enter the human genome from elsewhere all the time. For example, the book, Lateral DNA Transfer, reports, "Analysis of the draft sequence of the human genome reveals that is contains fully 8% endogenous retroviral sequences" (p 204). And these are not necessarily only viral genes, but may be other genes they carry.
With regard to our research project, 1) Is there any reportable result at this point?, and 2) What are the prospects going forward?
Best regards. Brig
Date: Tue, 05 Nov 2002 16:23:09 -0600
Dear Tom -- [...]
> Recall our poster at the astrobiology meeting? You show two graphs
In my concept, the mutations that could improve a genetic program would be capable of finding beneficial genes only if the genes were very nearby -- within an adjoining sequence space that has no more than 10^8 members, say. This would correspond to maybe a dozen nucleotide substitutions. That leaves us in the 98-99% similar range for an average gene. Still on the horizontal blue line. More likely, new genes for new functions have new sequences that are different from previous ones by more than 25%, or c. 250 nucleotides. Now we're down the slope on the red lines. But this sequence space (250 nucleotide changes) has some 10^150 members. I don't think the random search method will ever begin to explore it.
I realize that recombinations are also part of the process, but only a very tiny fraction of all possible recombinations are within the small-enough neighboring sequence space; so recombinations do not avoid the problem, in my view. [...]
I like your comment, "without having to generate custom data." Why can't we look for evidence of the uniquely human genes in the genome data that are already available? Isn't the chimp almost done? With just chimp you should get lots of data points to plot. This seems feasible to me. Your comments are invited!
Thanks. Sincerely, Brig
Date: Fri, 08 Nov 2002 12:19:45 -0600
> I don't think anyone believes that mutation has to explore the whole
(They used to think there was time available to do that.)
> But let me suggest a less theoretical approach.
This sounds different from our original plan, but also interesting. You seem to be stressing amino acids versus nucleotides now. To be sure of plausible mutation pathways, don't you have to look at nucleotides? If not, this might be a good approach.
In any case it would be important at the outset to state the expectations of darwinism and those of strong panspermia, and then to record all the results.
[...] Non-comparative studies will clearly reveal transfer if the flanking sequences are still present. Perhaps if you knew how rapidly these are lost, you could estimate all the examples of transfer. If you had stated the two sets of expectations at the beginning, than you could see which of them the extrapolation favored. Sounds complicated? Or maybe there's another way to analyze the data.[...]
Glad to be considering next steps with you. Best regards. Brig
Date: Thu, 19 Dec 2002 09:19:10 -0600
Tom, sorry to be slow. A lot is going on, and I have been pondering yours...
> ...Suppose that we
I think this approach might be interesting, but I do not see how it could be used to impartially seek evidence distinguishing between darwinism and strong panspermia. The most it could do, it seems, is either support darwinism to some degree -- a lot or a very little....I further think that lots of people are doing a similar thing already, and the results supporting darwin (new genes producing progress by darwinian evolution) are extremely weak. But your elaboration is welcome. I'm open-minded.
> The other thing would be to look for evidence of horizontal transfer.
This appeals to me, and I am already doing (to the extent I am able) the literature search, posting the results on my page about viruses. I think the momentum here is very strong in support of strong panspermia, but the same problem as above arises: this evidence by itself will not compare the differing predictions of the two theories.
BTW, I believe that looking for evidence of horizontal transfer is exactly what the people in John MacDonald's lab at the University of Georgia are doing now. (Incidentally, John made a remark to the student newspaper there that slightly bothered me. You especially might be entertained by the remark. See http://www.panspermia.org/whatsne26.htm#021128 and follow the internal link to "2002, 27 August: "Red and Black article")
My original goal was to find a respected researcher to undertake the question, "Is sustained evolutionary progress in a closed system possible?" Of course the relevance to biology is why this question is interesting. Hell, Tom, I don't know what to propose. I will help do the literature search, if that's the best thing. I will support the first idea, if I can see how the result might help distinguish between the two theories....
Tom, this has been a long one. Sorry. Thanks. Happy Holidays. Brig
Date: Wed, 22 Jan 2003 17:24:07 -0600
[Tom Ray wrote]
Dear Tom --
I would like to understand your logic. You said this was a brief answer. I may be ready for the lengthy answer. I had originally thought you could plot everything and see if the resulting picture is a network (supporting darwinism), or islands (supporting strong panspermia). You seem to be suggesting that we look at a much reduced group. I'm back at square one. I'm hoping for an impartial test, for which the two theories make different predictions -- we would state the predictions in advance and then do the test. But I am truly open to any good idea.
Date: Tue, 04 Mar 2003 23:15:16 +0000
I suspect that smart operators might be able to piece together pre-existing genetic programs that are acquired in pieces, over time perhaps. That would be entirely consistent with cosmic ancestry, and unexplained by darwinism. If so, evolutionary progress in a closed system would not extend beyond the piecing together of the pre-existing programs. However, if the smart operators can compose new programs, then, wow! That would be extremely interesting, and not at all what I am expecting. It would rescue darwinism spectacularly. (And in a way that darwinism has not predicted. How do the operators do it??) In the former case, sustained evolutionary progress in a closed system is not possible. In the latter case, it is. Regardless of my own inclination, I want to know what's really happening. If the operators can do what you are suggesting (that may lie down the slippery slope) I want to be the first to know about it! It would trash cosmic ancestry, but it would be a very important result....
Best regards. Brig Klyce
Date: Mon, 17 Mar 2003 14:59:25 -0600
Brig: [you wrote]
> Tom, this sounds very promising to me! I am glad that you are using 2-D
My idea is to begin by doing this analysis with only one species, human. I will be looking at the shapes of the distributions of gene families in sequence space. Suppose that we have 30,000 genes in 3000 gene families, in the human genome. On the average, each family has 10 genes. However, some families are very much larger with 1000 or more members.
There is no time axis in this analysis. We are taking a snapshot in time, of evolving gene families. We know that through evolution, duplication and divergence can increase the number of genes in a gene familily (within a species). But a gradualistic darwinian process would have different implications than a strong panspermic process, as to how genes in a gene family would be distributed in sequence space, at any point in time.
In the strong panspermic process, genes never change very much, so when a large gene family forms through duplication and divergence, all the points would tend to be tightly clustered in a spherical distribution in sequence space.
In the "strong" gradualistic darwinian process, changes in genes can gradually accumulate to the point that new gene families are formed. This implies that at any point in time, there are likely to be families with a wider dispersion in sequence space, and these clouds would tend to take-on non-spherical shapes as they spread into new functional domains. A spreading gene family that is in the process of forming a new, functionally distinct, gene family, would tend to take on an hour-glass shape, before completely splitting into two distinct, fully-separated gene families.
What do you think.
Date: Mon, 17 Mar 2003 23:14:54 +0000
I think the islands in strong panspermia can be any shape. But they will be isolated, like the Hawaiian Islands, or the lakes of Minnesota. Pieces that might appear to be budding off might not necessarily be on the way to new functions -- they might be on their way to extinction, for example.
The islands of darwinian evolution would look connected like the Aleutian Islands and the Alaskan Peninsula, or maybe like Japan, or like Indonesia. So it's not the shape but the connectedness between families known to have different functions.
BTW, by '"strong" gradualistic darwinian process' do you mean utilizing single nucleotide mutations (SNMs) only? To me, the critical question is related to the concept of Hamming distance --
1. How many single SNMs are required to go from gene A to gene B? Say the answer is g.
2. How many possible different genes could be created from Gene A by g SNMs? Say the answer is h.
3. Is the likely number of point mutations in A, since the separation began, on the order h? If so, the evolution of B from A by SNMs is not unreasonable. And this might be evident because the varying alleles will spread out into the sequence space enough to touch or overlap A with B.
And I think it might be possible to use the same reasoning with respect to recombinations within gene A (no imported sequences, just rearranged ones.) So more than SNMs could be analyzed this way.
However, for imported sequences, even if there is only one event needed to go from A to B, how do you count the number of possible importations? Aren't they too numerous to count, because they could
(This is related to my earlier question -- how would you map the Tre2 gene versus the two genes from which it apparently got its two halves?) [see Humanoid gene arose abruptly?, 18 February 2003.]
Finally, can we get buy-in from other darwinians on the premise that by mapping the gene families in a single species, evidence of evolution should be apparent?
I hope this is helpful.
Best regards. Brig.
Date: Mon, 31 Mar 2003 15:18:37 -0600
One thing we need to consider, is that under the Darwinian process, all islands start out as connected, but things are allowed to disappear (go extinct). This can cause things to become disconnected after some time. But there must be a different pattern resulting from a Darwinian process where everything starts out connected, as compared to a strong-panspermic process where everything starts out isolated. That is what I am trying to define.
Please take a look at: http://www.his.atr.jp/~ray/pubs/zen/node4.html where I discuss evolution in sequence space. This is a Darwinian analysis. Do you agree with it, or differ with it?
I mean no unconventional genetic operators. The conventional operators are mutation, insertion, deletion, duplication, inversion, cross-over and chromosomal rearrangements.
> Finally, can we get buy-in from other darwinians on the premise
I think so. We can probably find something on this in the literature.
Date: Tue, 01 Apr 2003 22:10:23 +0100
> Please take a look at: http://www.his.atr.jp/~ray/pubs/zen/node4.html
I agree that you have accurately stated the darwinian case. I disagree that it works that way. I think the "fitness gradients" are all very short and do not connect big clouds to each other. I really like the way you described the scene. Surely there is a way to quantify the image. Isn't that what Eigen tried to do? Except his fitness gradients were very short, right?
Date: Mon, 12 May 2003 14:40:07 +0100
This feature of high dimensional spaces is very interesting. Given that the Hamming distance between two completely unrelated average genes (1000 nucleotides) is only about 750 at most (25% of nucleotides will be the same), the existence of some gradual routes, with potholes, is perhaps not so surprising. But gravity (chaos) will still pull evolution downhill unless the upward slopes are wide and the rewards for incremental progress are powerful. I think.
BTW a peak is a good place, right? The only thing better is a higher peak? The place to avoid is the valleys (canyons, I suspect)?
I am starting to like better our historical reconstruction approach!
Date: Mon, 12 May 2003 16:20:18 +0100
[Tom wrote] > Anything higher is better, regardless of whether it is a peak or not.
Tom, when it is time to design a prize for lifelike evolutionary progress, it will be useful to say precisely what the vertical dimension is. If I had to say right now, I would say it measures the Least Algorithmic Complexity of the programs (not the nonsense). But I can already think of objections to that definition. For example, a program can do something useless or harmful.
Have you got any thoughts on the vertical dimension? Thanks. -- Brig Klyce
Date: Tue, 13 May 2003 09:27:32 -0500
[Tom wrote] > > Anything higher is better, regardless of whether it is a peak or not.
[Brig Wrote]> ...Have you got any thoughts on the vertical dimension?Our discussion of landscapes has been in the context of fitness, which is not necessarily correlated with complexity. So I see this question as a bit of a change of subject, and I just want to call attention to that.
For purposes of the prize you will want to reward some kind of evolution of complexity. I think that defining complexity is a can of worms. I've seen a lot of talk about it around the Santa Fe Institute, and my impression is that most of the definitions are rather theoretical, and would be essentially useless at the moment of actually putting a measure on a "real" system. A lot of them measure some statistical property that I doubt really correlates with meaningful complexity. In fact "meaning" is something that seems to escape these measures. Imagine trying to detect the difference between carefully constructed gibberish and meaningful language (as strings of characters, text). I believe that gibberish can have the same statistical properties as meaningful language. The same issue arises in trying to distinguish functional protein sequences, from random sequences.
When I got down to putting a number on the complexity of Tierra, so that I could make an objective test, I chose to use a simple count of the number of cell types in the organism. Three cell types is more complex than two cell types. This definition of complexity seems clear and workable, but it is peculiar to the specific system.
A measure of least algorithmic complexity might also capture the differences between the two and three cell type creatures, but I'm not sure. I haven't worked with least algorithmic complexity, so I don't have an emperical feel for it.
I realize that when you are putting your money on the table you would like to have a clear and objective definition in advance. I support you in that, yet I also feel that different measures might be appropriate in different systems. It would be ideal to have some kind of flexibility to facilitate that.
Date: Tue, 13 May 2003 19:52:07 +0100
Dear Tom --
Thanks for your reply about the fitness (adaptive?) landscapes. I would like to share it with Stan Franklin, one of ART's two advisory board members. I am having lunch with him today.
> Our discussion of landscapes has been in the context of fitness,
But we were also talking about the evolution of "EQU", about which Lenski et al. say, "Its performance is more complex than any other one- or two-input logic function...." In any case, complexity is at least related to the invention of new genetic programs, which is what I want to know about. [see Computer model evolves complex functions?, 11 May 2003.]
> When I got down to putting a number on the complexity of Tierra, so that
At one of the astrobiology conferences I attended someone promoted something like this as the best measure of biological complexity number of different cell types. Another person said the best measure was the number of different gene expression products (because genes can express more than one product, with alternative splicing for example.)
Alan Lightman called me this AM to say that Seth Lloyd and Gerry Sussman at MIT were waiting for my phone call to schedule a meeting to discuss a prize for an A-life demonstration of lifelike evolutionary progress and MIT's role in it (administrator?). Lloyd said he wouldn't participate without you! I will ask them for available dates from 15 June to 15 August for a (daylong?) meeting in Boston. Can you give me some dates too?
Date: Mon, 25 Aug 2003 15:34:17 -0500
In 1995, a group of scientists at Ohio State found that some introns possess a capability called "homing." Such introns can insert themselves into intronless versions of the that normally carry them, at the exact same position they would normally occupy in the version that has introns (10). This happens by a surprising mechanism - the RNA transcript of the intron is spliced directly into the DNA gene. Then by reverse transcription, the DNA complement of the RNA is manufactured and added to the opposite DNA strand (11). The phenomenon of homing by introns is not yet explained by any theory. But it makes sense, if the precise insertion of introns is part of the process whereby new genes are installed. As the scientists at Ohio State innocently comment, "...introns... may well be the key to determining where new information is transferred into a gene."
10. Steven Zimmerly, Huatao Guo, Phillip S. Perlman and Alan M. Lambowitz "Group II Intron Mobility Occurs by Target DNA-Primed Reverse Transcription" p 545-554 v 82 Cell. 25 August 1995.
11. Jian Yang, Steven Zimmerly, Peter Perlman and Alan M. Lambowitz "Efficient integration of an intron RNA into double-stranded DNA by reverse splicing" p 332-335 v 381 Nature. 23 May 1996.
Date: Mon, 25 Aug 2003 19:11:55 -0600
Dear Tom --
Nature of 21 Aug has two articles and a commentary about a subject we touched on on the phone: genetic hotspots. The abstract of the commentary concludes, "modifications at just a few developmental hotspots underlie 'parallel' evolutionary changes that occurred independently in different species."
Commentary: "Developmental biology: Hotspots for evolution" by MICHAEL K. RICHARDSON1 AND PAUL M. BRAKEFIELD, Nature v 424, pp 894-895 (21 August 2003):
The abstract of the second article concludes, "Our results indicate that some developmental regulators might preferentially accumulate evolutionary changes and that morphological parallelism might therefore be more common than previously appreciated."
Second article: "Regulatory evolution of shavenbaby/ovo underlies multiple cases of morphological parallelism" by ELIO SUCENA et al., Nature v 424, pp 935-938 (21 August 2003):
The first article: "Genetic mechanisms and constraints governing the evolution of correlated traits in drosophilid flies" by NICOLAS GOMPEL AND SEAN B. CARROLL, Nature v 424, pp 931-935 (21 August 2003):
The genetic changes are minor. But if evolution works like I think, and if a new program becomes available after speciation, darwinists might observe it as above.
Brig Klyce * Astrobiology Research Trust
Date: Thu, 02 Oct 2003 12:06:45 -0500
Dear Tom --
Thanks for your thoughtful reply. I will look for a chance to visit you there. The workshop idea is still being explored, feebly, for MIT. Any news from Seth Lloyd will be welcome.
Here's an article that might interest you:
It springs from an article in Science: Stuart, J. M., Segal, E. et al. "A gene coexpression network for global discovery of conserved genetic modules." Science 21 August 2003 which did not catch my eye.
Did you notice on my what'sNEW that Lawrence Livermore National Lab will do isotope testing of bacteria from the high atmosphere?
Date: Thu, 04 Dec 2003 09:01:22 -0600
Dear Evan -- I saw that you wrote the introductory piece for Current Opinion in Genetics & Development Vol. 13, No. 6, December 2003. The whole issue looks interesting. I have poked around to find a copy but I am only offered a year's subscription. If you have any extras, may I request one?
In any case, I hope all's well with you. Any news is welcome. My latest preoccupation is testing of bacteria recovered in the high atmosphere to see if they are earthly or not. Isotope ratios, especially Nitrogen, could definitively show if they are from space. Lawrence Livermore has agreed to measure the ratios in January.
Best regards. Merry Christmas. Brig
Date: Mon, 09 Feb 2004 09:31:56 -0600
Dear Tom --
Thanks for your recent emails. I think I do not care to meet the creationist professor. I have made overtures to creationists elsewhere, including at U of Memphis, and my experience is always the same. They are as committed as darwinists to restricting the number of possibilities to two. But if there is ever a debate or forum between his side and some darwinists, I would like very much to represent a third point of view.
Thanks also for your earlier one asking me to tell you my objections to Long et al. I think it is a good idea, but a mishap around here puts me in an awkward position. I have lost the copy that Long himself mailed me. Over the weekend I did a major re-arranging of office furniture, and everything got put somewhere new. An hour of hunting has failed to turn it up. Could you email me a pdf? If you think any of the other 4 that your students are reading would interest me, I'd welcome it as well. I will then try to make a careful statement for you to consider.
Thanks. Best regards to you and Chenmei. Brig
Date: Tue, 10 Feb 2004 15:54:14 -0600
Dear Tom --
attached is a .doc with some of my complaints about Long et al. Because I think that "origin" is only one possibility, I think the evidence for it needs to be very strong, not merely almost plausible.
Date: Sun, 22 Feb 2004 17:22:27 -0600 (CST)
I have no doubt that HGT is overwhelmingly important in prokaryotes. I have no doubt that it occurs in eukaryotes. I doubt that it is an important force in eukaryotes. But you make a very much stronger claim.
You claim not just that it is important in eukaryotes. You claim that all progress, or at least all sustained progress in eukaryote lineages is derived from HGT. That is an extremely strong claim and I haven't seen evidence to support it.
I'm willing, even eager, to take a hard look at the evidence, and to generate fresh unbiased evidence where appropriate, in order to get to the bottom of the matter. I don't have a "preferred" outcome. I just want to know the truth, and I'm open to the possibility that you are right.
Date: Sun, 22 Feb 2004 20:15:21 -0600
I say: Darwinians claim that all progress is derived from self-sufficient darwinian processes. That is an extremely strong claim and I haven't seen evidence to support it.
Date: Sun, 22 Feb 2004 22:36:34 -0600
Tom -- You said that transposons act within genomes only. Here's a paper by good authorities that seems to say otherwise; a transposon is a transposable element, isn't it?
Margaret G. Kidwell and Damon R. Lisch, "Transposable elements as sources of variation in animals and plants," p 7704-7711 v 94 Proc. Natl. Acad. Sci. USA, July 1997. http://www.pnas.org/cgi/content/full/94/15/7704
[here's a section from the paper]: "TEs as Genomic Parasites" --
"The intrinsically parasitic nature of active TEs (41-43) accounts for their undisputed ability to invade new species, increase in copy number, and survive over long periods of evolutionary time. The replicative advantage of TEs (44) is responsible for this ability, which is facilitated by their generally compact structure and inclusion of the coding capacity for transposition within their sequences. Natural selection acting on TEs at the level of the DNA sequence is responsible for maintaining their essentially parasitic properties. For example, P elements can rapidly invade a naive population of D. melanogaster, despite extremely strong negative selection at the host level in the form of high frequencies of temperature-dependent gonadal sterility (45).
"A proclivity for horizontal transfer is consistent with the role of TEs as genomic parasites. The life cycle of TEs in any single phylogenetic lineage can apparently last for many thousands or millions of years and can be considered as a succession of three phases: dynamic replication, inactivation, and degradation (46, 47). The transposition of both major classes of elements is error-prone and produces nonautonomous elements that often repress the transposition rate of active elements. Over long periods of evolutionary time, there is a tendency for a family of elements to degrade in coding capacity, but horizontal transfer to another host lineage provides the opportunity for active TEs to move to another lineage and begin the cycle over again (46, 48, 49). There is evidence that TEs do transfer horizontally more frequently than nonmobile genes (46). Class II elements such as mariner and P elements provide good examples of TE horizontal transfer (50-52), but a major puzzle remains regarding the mechanism by which horizontal transfer is achieved (49)."
-- Brig Klyce
Date: Sun, 22 Feb 2004 23:07:56 -0600
"Hundreds of human genes appear likely to have resulted from horizontal transfer from bacteria at some point in the vertebrate lineage. Dozens of genes appear to have been derived from transposable elements.
"Although about half of the human genome derives from transposable elements, there has been a marked decline in the overall activity of such elements in the hominid lineage. DNA transposons appear to have become completely inactive and long-terminal repeat (LTR) retroposons may also have done so.
"To survive, DNA transposons must eventually move by horizontal transfer to virgin genomes, and there is considerable evidence for such transfer (149-153)."
These are some selected sentences from: "Initial sequencing and analysis of the human genome" by the International Human Genome Sequencing Consortium, Nature 409, 860 - 921 (2001)
Sat, 31 Dec 2005, 10:42 AM CST
Fri, 15 Sep 2006 11:53:06 CDT