Sunday, March 10, 2013

DNA, Computer Programs, and God

Is life designed? Or is it the product of chance and natural processes? That is the topic of this blog.

The motivation for writing is to sharpen the idea I introduced in a previous blog, which went viral with more than 15 page views (some of which, I admit, may have been mine).

That question was can random addition of information (which is the premise of evolution) ever result in an improved version. My premise was that it would not. I used as an example the addition of random letters to a book. In my test of the proposition, no useful information was added. My conclusion was that no matter how many times the process is attempted no new useful information would be added. That seems an intuitive conclusion. But a friend pointed out that I had stacked the deck. I had created a strawman argument. And he was right; I had oversimplified. This blog is an attempt to test my premise a little more realistically.

First, it is important to understand what means and processes drive evolution. They are mutation, random genetic drift, and gene flow [1]. There is, however, the probability that additional forces and processes play a part[2]. Natural selection then acts upon any newly expressed features, selecting out those organisms less viable or less able to pass on their genes to the next generation.

In simple terms (if this isn't already too simple) any new feature, such as eye color, arising in an individual by mutation must be passed on to the next generation if it is to become part of the population and result in any change in the succeeding generations. This is actually unlikely to happen because the cell has a self-correcting mechanism that usually eliminates mutation. One self-correcting process is called recombination. Recombination assists in DNA repair writes Suzanne Clancy, Ph.D on this page [3]. But let's imagine that the mutation is passed on.

For example, by a genetic mutation a single plant may have curled leaves. That gene might be passed on to the next generation of that plant. [A].

If the new feature is passed on, curled leaves may give that second generation plant a selective advantage, meaning the curled leaves enable the new curled leaf plant to survive and reproduce better than the plants without curled leaves. That is evolution in the most limited sense. Or curled leaves may give the second generation plant a selective disadvantage. In that case, the new feature and the mutated gene that caused it will pass out of the population, and no evolutionary changes occurs.

Fundamental to all of this is DNA. DNA is the information (instruction book) that drives the development of every living organism [4]. DNA is is every cell, including those cells which will be part of the reproduction of another organism.

Now, is there anything like DNA with which we might compare and test the processes of evolution? It turns out there is. A computer program is very much like DNA. If I may quote Bill Gates, "DNA is like a computer program but far, far more advanced than any software ever created" (The Road Ahead, p. 228). But many others have noted the similarity as well. This article [5] on quotes biologist Leroy Hood and biotech guru David Galas and MIT Professor of Mechanical Engineering Seth Lloyd.

If DNA is very much like the digital code of a computer program, does it function in the cells as a program? Yes, says Bill Gates, among many others including Glyn Moody in Digital Code of Life, p. 4,5. Then the question is can a computer program differentiate (change) by the same means suggested for living organisms, that is by random unguided processes. That would seem to be an interesting test of the theory of evolution.

Now, programs can be exceedingly simple or exceptionally complex. My first attempt at programming at was >alert ("This is my first program."). It created a box with the message in it. But obviously most programs are a bit more complex. For example, the code for a simple E-coli bacterium contains 1.5MB of information. The human genome contains 1.5GB + of information. The program for E-coli is small as programs go. The program for humans is rather large.

So, could a computer program function if something like a point mutation occurred? A point mutation would be a change in one letter of a line of code. If that letter or symbol were the > in my small program, the program would not work. If that letter were a letter in the message, only the message would be changed, and then only by a very small amount. My brother-in-law, an experienced programmer, told me of another example.

I once troubleshot a problem when an inadvertent energy pulse would take command of an address/data bus and zero out the information contained on the bus at that instant. The resulting set of zeros became a hex command (00) and a data byte (0000) which translated into a program that reset the 'seconds' register in the real time clock. This 'mutation' is a perfect example of either a random functional change to an existing software organism or a simple random 'standalone', ' functional' software organism.
So, it is possible for a point mutation in a computer program to create functional change. How about in a living organism?

In a program the size of the genome of E-Coli bacteria, 1.5MB, it is unlikely but possible that a point mutation would cause a serious problem. But it is also unlikely that a point mutation would change any functionality or result in a new (novel) feature in the program, one that would separate the program with no mutation from the program with the mutation. My brother-in-law's example is, I think, rare and noticed only when it does create a problem or by computer programmers, whose job it is to maintain the functionality of the program. Most of us with computers live with various problems due to mutated programs without serious complaint. There is some similarity to living things.

In living things a point mutation is rare, particularly in the reproductive cells, which is the only place it would matter in evolution; and it is most often neutral in effect. It is silent, meaning it causes no change in the features of the organism.[6] Sometimes, however a point mutation results in the death of the organism's offspring. Much more rarely (to the point of being exceedingly rare) is there any result that is beneficial to the organism, and often those mutations simply "allow genetic diversity to exist within a population, increasing the range of alleles."[7]. Alleles are alternative forms of a gene. One allele will produce blue eyes, another brown. A population, such as the population of all humans, is made more resilient when there are more alleles. So in this way a population is benefited. But point mutations do not result in evolution in the larger sense of speciation [B] or the addition of information to the genome.

A frame-shift mutation has a greater effect and usually makes the DNA sequence meaningless [8]. Because there is redundancy in the genome, a frame shift mutation may not result in the death of the the offspring, though often it does; but it does not result in improved functionality either because it subtracts useful information. I repeat, most mutations are neutral in the effects they have on the organism [9]. In a computer program a frame-shift mutation would be similar to the scrambling of line of code. Depending on how crucial that line is to the program's function the scrambling of a line of code may mean nothing, or it may crash the program. It will never improve functionality.

Another class of mutation is gene duplication. In a computer program that would be the duplication of a line of code. That does usually have an effect on the functionality of the program, according to my programmer daughter, but it probably won't crash the program, though it could. Usually it duplicates a function. What it does not do, unless it is purposely done by the programmer, is create new and beneficial functionality in the program. (Has anyone reading this ever turned on their computer in the morning and found that any program has become more functional overnight by a random "update?")

In living things the results are similar, though duplicating a function in living things often causes weird additions that make the organism non-viable. Lab experiments have caused these kinds of mutations in organisms such as fruit flies. The result may be another set of wings or eyes. Usually these duplicate features are useless, however. And useless features are deleted out of the population by natural selection.

My conclusion regarding the possibility of random mutations providing new information and thereby increasing the functionality of either a computer program or an organism is that the possibility is so rare as to be non-existent, though there is actually more chance of this happening in living things than in a computer program. Those mutations, however, are not evolution. They create diversity, but diversity does not necessarily result in new species.[10]. In the programmer's world a new species is always the result of purpose and design. Is it also so in the natural world?

But what about genetic drift? Could it provide the means of natural unguided evolution?

Genetic drift is the random change in the frequency in which any one allele (a gene for one variation of a feature, such as blue eyes) is found in the next generation of a population. If in one generation there are ten blue eyed people and ten brown eyed people and in the next there are four blue eyed individuals and six brown eyed individuals, genetic drift has occurred. The result over time may be that brown eyed individuals continue to increase in frequency. Eventually, there may be no blue eyed individuals. Evolution has occurred, according to the broadest definition of the term. But notice that no new feature has arisen from genetic change, only the frequency of already existing traits. And no additional information has been added to the genome.

In computer programs that may be the equivalent of waking up in the morning, turning on your computer, and finding that the theme on your browser was different. You had a theme on the browser when you turned off the computer. Now it is different, but it is not a novel feature and it provides no more functionality that the older version.

Finally, there is gene flow. Gene flow is a lot like genetic drift. It changes the frequency of one variation or another in a population. It does so by mixing individuals from one population with another. In biology gene flow actually works against evolution by reducing the amount of differences between one sub-population and another.

In computer programs it might be the equivalent of in open source program such as the browser I use, SeaMonkey, where a programmer may add a variation, such as the theme color of the browser. The variations may keep users entertained enough by the browser to keep us from changing to Chrome, but in this analogy, no additional information is added. The only thing that has happened is that the information in one line of code is changed. Perhaps in time one color theme dominates and another color is discontinued. But again, no additional information is added.

Programmers, forgive me. My understanding of programming is limited. However, the real issue is biology not programming.

The bottom line is that most of the means and processes that are suggested as the drivers of evolution do not - by themselves - any more than random mutations, or the frequency of variations in a computer program change a program in any significant way, especially in any beneficial way.

That begs the question why so many biologists, who know in far greater detail the limits of these means and processes than I, believe that evolution (in the larger sense of the word) is a purposeless natural process. It seems so unlikely. I venture to guess. First, there is no other competing "scientific" theory. Now, I can understand believing that change over time is self-evident. I do not understand why that would lead to the conclusion that change is mindless. Remember, the genome has increased from 1.5MB or 1.5GB. That is big. Virtually all the information in the human genome is vital to the functioning of our bodies. Evolution must account for the increase. I don't think it does.

The second is that for some, adding a director violates Occam's Razor. Occam's Razor is the principle in logic that the simpler explanation is better unless the more complex explanation has greater ability to explain the phenomena observed. But the fact is, natural unguided evolution does not explain the phenomenon of the diversity in living things we observe. Adding an intelligent director does.

Finally, it is possible that the issue is metaphysical rather than scientific, or perhaps in modern parlance, a matter of worldview. A worldview is what we believe to be true. Though a worldview is developed over time, it is largely the result not of facts but of suppositions acquired from our culture and family. My worldview includes a Creator. Another person's worldview might be wholly materialistic allowing no room for a Creator. My default conclusion is that the facts add up to the need of a Director. A materialist's default conclusion is that the facts add up to no need of a Director.

In my opinion, if the issues are science and logic, the reasonable conclusion is that the natural world and the change that has occurred over time argues for the need of a Director just as a new version of a computer program argues for a programmer. If the issue is worldview,however, facts and logic don't persuade - and that is true for me as much as it is for the materialist. I do not believe anyone is neutral here. The argument that I trust only the facts is specious. Everyone interprets the facts, and interpretation is done in the context of worldview.

For my part, I find science and logic support my conviction that the living world must have a Director/Designer. If I look beyond the living world to the cosmos, the subject of the next blog, my conviction only grows stronger.

No comments: