In What Way are Mutations Random

This thread is meant to find agreement on what the scientific definition of random mutations is within the domain of science and the philosophical boundaries that this definition resides within. Random mutations are one of those topics where misunderstandings are common, both from a scientific and theological point of view. In fact, my understanding of these concepts could be wrong, so I encourage feedback from biologists and laypeople alike. I am hoping that we can have a productive conversation that seeks to find common ground (as much hope as one can have on an internet discussion forum :wink: ).

From my understanding, random mutations are random with respect to fitness. This means there are no scientifically established general mechanisms whereby the organism can sense an environmental challenge and mutate a specific base in a specific gene to overcome that challenge. Instead, the processes that produce mutations appear to be blind to the needs of the organism. This is exemplified in the plate replica experiment described by Joshua and Ester Lederberg in their paper from 1952:

In that experiment they started with a single bacterium which then went on to found all the populations they used in the experiment. This is important to note because if the mutation were present in that founding ancestor then it would be present in nearly all of the descendants. Without going into the specific of the experiment, what they observed is that some of the bacteria in the experiment were resistant to antibiotics, a trait not found in the original ancestor. What they also found was that the mutation occurred without the bacteria being exposed to antibiotics. In other words, the mutation wasn’t caused by exposure to antibiotics and did not appear to be affected by the needs of the bacteria. The same conclusions were reached by Luria and Delbruck in their fluctuation assay in 1943.

I mention the dates of these papers, 1952 and 1943, for a reason. The structure of DNA was discovered in 1953, a year after the Lederbergs concluded that mutations were random with respect to fitness and a decade after Luria and Delbruck’s work. Random was not defined as being an even distribution of base changes along a stretch of DNA. Random was not defined as all substitutions being equally probable. Nobody even understood what mutations looked like at a molecular level at that time, but they still concluded that mutations were random with respect to fitness because it had to do with the relationship between mutations and the environment.

There are also exceptions that prove the rule. CRISPR/Cas9 is a really good example. In this system, the bacteria specifically alter their DNA to acquire immunity to specific phage. However, this mechanism is very limited in scope and can’t explain the vast, vast majority of mutations that occur throughout biology.

These are all scientific conclusions, and it isn’t meant to be a Truth with a capital “T”. Science is necessarily limited in its philosophical scope. All science can say is that there is no statistical signal linking mutations to the needs of the organism. This scientific concept and guidance by God can easily coexist, as many on this site have discussed. There is no way that random can be construed to mean no God or no guidance by God.

Please correct any science errors you think I may have made, and please try to find any common ground you think exists. Thanks!


Oh yes there is.

Perhaps in a philosophical sense, but not in a scientific sense.

My favorite mutation is lactose tolerance.

I hope to learn more from this thread.

1 Like

Feel free to ask any questions you may have.

1 Like

To some degree it will always be blind.

On the other hand, more complex organisms tend to exert all kinds of controls to decrease the likelihood that mutations will be harmful. It may be purely random in the case of viruses but even bacteria have means to limit mutations to more productive areas of the genetic code.

They are the same. They meet in rationality.

How, where.

What makes the subject confusing is that ‘random mutation’ means two overlapping things, I think. One is the notion described by @T_aquaticus; the other is the more usual scientific meaning of ‘random’, which is that we can’t predict which mutations are going to occur except by a probability distribution.

That’s the formulation I go with. ‘Random with respect to fitness’ kind of captures that, but also isn’t strictly true, since mutations are biased against being deleterious. (That is, the kinds of mutations that occur the most frequently are less likely to be deleterious.)


Being biased against deleterious still isn’t responding to a new specific challenge, which is how it sounded to me that @T_aquaticus was stating it. It is more of an advantage learned with respect to previous challenges, adapting to the fact that variations in some sectors of code were more harmful than helpful in the past. Though… in the long run I suppose this could go in reverse and reopen those sections of code should variation there become advantageous. And perhaps in some sense that adds up to the objection you were making.

For those unfamiliar with mutagenesis basics… In organisms more complex than viruses, most damage to genetic material is simply repaired. Bacteria (E-coli) have been found to use UV radiation for introducing variation into their genome by selectively protecting damage from UV radiation from their own repair process. But other organisms have completely different ways of introducing variation into their genome like sexual reproduction, chromosome crossover, and more. Since these are evolved mechanisms rather than random damage from environmental causes, the limitation to less deleterious variations is pretty obvious.

Ah. Thanks for that. Link?

Definitions in biology tend to be imperfect generalizations.

Are there any good references that address how mutational biases protect against deleterious mutations?

My understanding is that these types of mechanisms act on the whole genome, not just in functional regions. Of course, the majority of bacterial genomes are functional so it may be a distinction without a difference in that case.

Here is a scientific paper which finds that some E-coli even count how much damage they are getting from UV to regulate the use of an error prone polymerase which promotes mutagenesis.

The explanation I gave before was from a textbook on mutagenesis which I found and read in the university library many years ago. In that text it reported that E-coli had a molecule which actually covered and protected damaged sections of its code from its own repair mechanism.

Of course, glipsnort is the expert on this topic here. As a physicist, I am just one of the dabblers.


Are you able to clarify for me…

Do you mean “kinds of mutations” in the sense that point substitutions may be more or less frequent than frame shifts?

Or do you mean e.g., that certain point mutations along certain strands of the DNA where such mutation could be significantly deleterious are less likely than mutations in other locations where those point mutations might be less deleterious?

I mean that point mutations that are transitions happen more readily than transversions, and are also more likely to result in the same amino acid.


On the transition/transversion front, there’s this paper. More broadly, there’s this. I haven’t seen anything on the latter topic recently (not that I’ve been looking).


This would be an example of increasing the random mutation rate through the expression of a polymerase that makes more random mistakes.

The SOS response also has an interesting place in the history of random mutations. If memory serves, they saw an increase in beneficial mutations after UV irradiation (lac+ revertants on lactose plates). The evidence started pointing towards non-random mutations which made them investigate the mechanisms that were in play. What they found is the SOS response, and instead of non-random mutations they found the genes responsible for an increase in the random mutation rate.

With respect to fitness, these mechanisms are just as likely to repair a beneficial or neutral mutation as they are a deleterious one.

Another interesting mechanism is the increase in mutation rate in actively transcribed DNA within bacterial genomes. It is thought that single stranded DNA might be more susceptible to mutation than double stranded DNA.

Yes, they have no way of distinguishing whether a mutation will be beneficial or not. BUT what they can do is protect crucial areas of code where almost all mutations are disastrous.

For those who think God is just as much in charge of genomes in an Old Universe scenario as He would be in a Young Universe scenario - - every mutation might LOOK random to human understanding… but God is in charge of ALL mutations.

“Let your conversation be always full of grace, seasoned with salt, so that you may know how to answer everyone.” -Colossians 4:6

This is a place for gracious dialogue about science and faith. Please read our FAQ/Guidelines before posting.