Can random mutations cause an increase in information in the genome?

One of the problems here derives from how we choose to define 'information'. In a mathematical definition, increases in information are merely a statistical property of the encoding system, so genes with more base pairs contain more information than those with fewer, just as sentences with more words theoretically have the potential to contain more information. But, as we know, there are a lot of very long sentences written that contain less information (in a qualitative sense) than some very short sentences.Another problem with discussing 'information increase' is one of context. Your example about a gene with eight alleles refers to information contained in a *population* - not an individual, as any given diploid individual can only have a maximum of two of them.But the clear answer to the topic question is an undeniable and resounding YES and it is best explained with a genetic illustration.Suppose a random mutation next to a functional gene causes it to be duplicated during meiosis so that all germ cells arising from this lineage contain two (tandem) copies of the same gene. However, only the original gene is transcribed in the organism - the duplicate lies dormant for many generations because its promotor region is suppressed. During this period of dormancy, more random mutations accumulate in the duplicate gene that are not selected against because the is no expression and the amino acid sequences encoded by the gene are completely changed. Genetic information in the organism has been increased - but it will have no consequences to organismal phenotype unless another random mutation activates the promotor of the duplicate gene that is then transcribed into a completely novel protein. At this point, protein information in the organism will have been increased.It does not run counter to any logic to assume that random changes can result in increases in genetic or protein information. What is required is an understanding that these random changes occur within the framework of regulatory mechansims that are highly structured and conserved, so that some random changes are conserved (either through positive selection for them or simple neutrality of effect) while many others are not (deleterious - selected against).
Thus, even though the original event is truly random, it can result in increased information because neither its conservation nor its loss will be random events, but rather directed by very complex and conserved biological processes.

The problem is there are probably as many ways to measure 'specified complexity' as there are possible definitions of it. So my inherent preference is for concrete, biological examples, rather than for quantitative mathematical techniques that assess only the 'potential' information encoded in particular systems.You are right that, from an evolutionary standpoint, information contained in the population is the most important. We can see the important effects of information loss at this level when populations pass through a 'bottleneck' - substantial genetic diversity is lost to the point where survival of the species may be endangered. So population information content is always some function of population size.The argument for information increase as a function of random mutation is actually very easily made here. Random mutations occur at some statistical rate in the population (bigger populations will have more variation) and UNLESS every single one of them is lethal or sufficiently deleterious as to be imediately eliminated, the amount of information in the population *has* to increase over time, measured in number of generations. (Remember that many mutations are effectively neutral to fitness, even if most are deleterious and very few are beneficial).Examples at the individual level are, as you note, more relevant to the average person, who often needs reminding that evolution does not occur at the level of the individual. But we can provide examples at this level as well. Th simnplest way we might quantify information contained in individual genotypes is by degree of heterozygosity. For example, a mutation in a transcribed (functional) gene leads to formation of a unique and novel allele at one locus in a germ cell during meiosis. The zygote formed by this germ cell, being diploid, has two copies of the gene, one being the novel allele. This individual, instead of producing only one type of protein product, produces two slightly different ones, and contains more 'information' (at both the genetic and protein levels) than other individuals in the population that are all homozygous for the ancestral allele. Voila - random mutation produces an increase in information at the level of the individual.This message has been edited by EZscience, 02-15-2006 10:41 AM

I think that anyone can 'see' intelligent design almost anywhere they want to in living things (and elsewhere), but in reference to your quotation, I would argue that random mutation generates the (potential) complexity of living organisms, but the highly ordered mechanisms of inheritence selectively preserve portions of it in very specific biological contexts. Hence the illusion of directed processes and/or designed systems when the end results are viewed.

All this arises because 'specified complexity' is not a scientific concept in any way, however Dembski may contrive to make it appear as such.The anti-evolutionists love to phrase arguments in terms that are neither quantifiable nor testable.I think PaulK has a good point. Dembski is trying to define something that is 'a priori' impossible to produce via evolution. We should not be suckered into trying to grapple with his contrived concept.

Quite so.
But I would contend that the topic sentence of the opening post is answerable and was never in contention.
Random mutation generates genetic variation that, acted on by ordered biological processes, clearly produces changes in living structures. Life's complexity is ordered by natural selection and 'specified' by the molecular rules of genetic expression.What I can't understand is why these ID people refuse to construe such genetic changes (which they accept) as comprising 'new information' or 'increases in information content' both of which they clearly are.

The problem is that 'information content', by any primitive measure, does not equate to 'organismal complexity'. If it did, then more complex organisms would have the largest genomes. They don't. If you go here you will see the absence of relationship between the two depicted graphically.Larger genomes have the *potential* to encode more information, but that information may or may not exist (much of the genome may be random repeats), or the same amount of genetic information may encode either simple or complex structures.
Along these lines, a single gene may cue a whole cascade of developmental events leading to a very complex change in morphology (= huge increase in complexity) with a very minor change in 'information content' of the genome. Thus I would argue that
'information' is not a useful variable to try and quantify in biological systems. Information content doesn't evolve in populations; gene frequencies and their mechanisms of expression and inheritance do.

Yes, the argument is ridiculous to anyone who understands the genetics of inheritance. But you raise another question. How can we measure biological complexity in an objective way?
We can't.
We would have to select a series of quantifiable criteria and these would necessarily be susceptible to intrinsic bias.For example, one criteria might be 'number of transcribed genes', but this would be biased toward ranking genetic complexity more highly than organismal complexity and, as I have already mentioned, these are not the same things.Criteria for organismal complexity are even more difficult to establish because not all morphological features are shared by all organisms. If we counted appendages, we would rank jellyfish higher than humans. Similarly, if we measured behavioral complexity as, say, the number of distinct intra-specific signals the organism is capable of producing and responding to, we bias our measurements toward organisms that are behaviorally complex, and against those that might be more complex in morphological ways.Complexity is not quantifiable in any way that will be objective and equally applicable for all living things. Degrees of complexity are not a focus for biologists. Thats why we don't have teams of biologists trying to re-organize Aristotle's Escala Naturae according to levels of organismal complexity - rather, we have abandoned it entirely.Complexity in biology is kind of like pornography - it lacks a clear and uniform definition, but you know it when you see it.So the answer to your question is likely dependent on some sort of anthropic principle.If you are a dog owner - you will judge the dog more complex.
If you are a cat owner - you will judge the cat more complex.If you own both (as I do) you will realize that there is no fair answer to the question.ABE: This not to say we cannot objectively determine whether a hamster or a human is more closely related to a dog or a cat, because this we can do.This message has been edited by EZscience, 02-15-2006 01:56 PM

Generally quite correct. But on the subject of complexity,
one small qualification is in order.It seems to you that the general tendency of evolution over the long scale has been to increase complexity of living things.
This is generally true when we consider higher organisms and those that dominate their ecosystems.
This is certainly true at the level of communities, as more species make for more complex interactions etc.
However, at the organismal level, it is important to note that increasing complexity is not an inevitable outcome even over long evolutionary time scales.
Consider the Archaebacteria, some of the oldest life-forms on earth that evolved in a reducing atmosphere and are now narrowly confined to anaerobic environments. Their success in surviving this long is likely their *lack* of complexity. They have not needed to evolve more complexity in order to maintain dominance of their particular niches, even if these niches have been reduced to obscurity in many cases, e.g. the guts of cockroaches and under-sea volcanic vents.Under some circumstances, complexity can be selected against. For example, mass extinctions seem to selectively eliminate the dominant, more complex animals, the survivors having quantitatively and qualitatively lower resource requirements for growth and reproduction.

It always works like a ratchet in the sense that there is no 'going back', no reversing the arrow of time.It is only that increasing complexity is not necessarily an outcome of this process. A species can be 'ratcheted' toward greater simplicity of body form if this improves survival in a particular niche. Whales evolved a simplified body plan relative to their bear-like ancestors because this was body plan was more conducive to remaining in a marine environment.

Not at all. Simple organisms necessarily preceded more complex ones. I am only pointing out that not all lineages are destined for increased complexity. Increasing complexity is not an *inevitable* outcome of organismal evolution, although where it has occurred, evolution is responsible. I am refering to complexity in specific lineages - some will increase in complexity - others not.
However, as the number of species in a community increases, the complexity of the ecosystem must necessarily increase. This is not to say that some organisms in that community will remain in very simple forms while others increase in complexity.But I really dislike this word 'complexity' because it has no real biological meaning. It is a subjective quality we tend to assign living things based on our own perceptions of them. It is not quanitifiable in any way, as I explained below.This message has been edited by EZscience, 02-15-2006 02:51 PM

The problem is you are equating 'change' with 'degradation'.
You also begin with the mistaken assumption that all gene products are perfectly functional or somehow 'optimal' to begin with. Not so. Functionality is very often determined by the genetic and external environment in which the gene is expressed. Plus some mutations produce sort of 'effective alternatives' to the original product. They are effectively 'neutral' under most conditions, but may provide advantages under others. So many genetic mutations may accumulate, and many alternative alleles may spread in a population, without necessarily having big effects on fitness. That's why every population is genetically heterogenous - even asexual ones.

Exactly. And why is this not logical?
Could it be you mistakenly assumed that evolutionary theory posited that all lineages necessarily increased in complexity over time? Because it happens to exploit a niche in which simplicity is favored.
Do you think that bird flu would be able to spread around the world and cross-infect birds and humans if it was as complex as an elephant?

Possibly. But I think this approach would still produce a bias toward morphological 'complexity' at the expense of behavioral. Furthermore, different levels of behavioral complexity could be achieved without different levels of cell differentiation. Your system would ignore differing levels of organization within cell types. Say two organisms each have only two types of cells in their brain, nerve cells and glial cells - they would get the same rating. But if one has a larger brain OR simply more complex arrangements of the same cell types, you would be ignoring this super-cellular level of organizational 'complexity' (Damn, I hate this word).There continues to be no definitive biological definition of biological 'complexity'. It means different things to different people. It can be measured, qualified, or quantified in countless ways, none of which will be definitively all-encompassing.On your other point, just because ID'ers and creationists are always hand-waving about complexity does not mean it needs defining. That burden falls on those who would introduce the term and bandy it about. In fact, their inability to objectively define 'complexity' is a good excuse for the rest of us not to wallow in intellectual mud trying to grapple with their flawed concepts. I am of the opinion we must insist on precise scientific terminology whenever we want to have a meaningful discussion of a scientific topic. And you can't have a meaningful scientific discussion about such a nebulous concept.No further comments from me until later today - a dental appointment approaches...This message has been edited by EZscience, 02-16-2006 09:10 AM

I guess I have no objection to the use of 'complexity' in relative terms when differences are obvious. A horse is more complex than bacteria. Tropical communities are more complex than temperate ones.I think it is the whole issue of 'absolute' levels of complexity that I really oppose. For example, if Dembski and his cronies are going to argue for 'specified complexity' as evidence of ID, they must come up with some absolute, quantifiable level of biological complexity that qualifies something to be considered specified as opposed to non-specified, and distinguishes something designed from something evolved. They are implicitly assuming absolute levels of complexity that do not exist and cannot be 'specified'. Just like the invisible boundaries that supposedly preserve biblical 'kinds' of organisms from macroevolutionary change.

Here is my summary of thoughts from this thread.Multiple lines of evidence indicate that genomic information can increase in lineages through a variety of well-understood genetic mechanisms, e.g. gene duplication and chromosome doubling to name two.Acumulatied increases in genetic information over evolutionary time scales are necessary to explain adaptive radiation of lineages that, in many cases, have generated increasingly 'complex' organisms. However, increasing complexity is not an inevitable outcome of increases in genomic information, although the latter is likely a requirement for the former. Nor do lineages surviving long periods in evolutionary time necessarily give rise to complex descendents. Some things are selected to remain simple.Finally, we see the 'invisible barrier' to macroevolution once again being raised by the anti-evo folks that would claim organismal divergence above the level of species cannot occur. I suggest that the real theoretical challenge here is to explain why two separate species (or whatever higher taxa you choose) should NOT become increasingly different in time. Any two populations that do not share a common gene pool anymore are genetically isolated and would be expected to diverge over time *simply by chance*, although the odds are they will also experience divergent selective forces unless they lead almost identical life histories in identical habitats (e.g. the special case of sibling species).The one constant in biological and ecological systems is that things change over time, whether this involves organisms aquiring new genetic information, or simply altering the expression of their existing genomes. The onus is on anyone who wishes to posit limits to the extent of that change to produce some hypothetical mechanism that might prevent it. If bacteria can change genetically to become resistant to antibiotics, what mechanism is it that prevents higher taxa from diverging into forms that are completely dissimilar from one another and their common ancestors?