I'm designing a creature that has found a new way to evolve directly during its life time. This method is different from having to bear incredible environmental stress or being a victim to violent radiation.

This entity strives to survive by stealing the genetic code of other organisms around and using it to adapt to new menaces.

The creature decodes the genetic code of something else, copying the useful parts, and then importing them into a compatible format like DNA to XNA and vice versa.

I imagine this ability could be useful for a myriad of things like fighting cancer, by simply stealing the genetic code of animals that have already become immune to it (like naked mole rats) or defeating a virus by using the genetic code of said virus for its own advantage.

Much like water bears, but with the bonus of choosing what genetic code to steal instead of letting fate decide.

Over time using the genetic code of other organisms can transform this entity into a completely different creature (without losing the genetic thief ability).

Is this any realistic, or is it impossible even for alien life?

  • 1
    $\begingroup$ Sounds cool, but how would they know the future ramifications of particular sections of DNA? $\endgroup$ Aug 12, 2016 at 22:46
  • $\begingroup$ A creature like that exists in Scratch Monkey by Charles Stross. $\endgroup$ Aug 13, 2016 at 14:11
  • 1
    $\begingroup$ The Zerg in Starcraft are kinda doing this. $\endgroup$
    – Skye
    Aug 16, 2016 at 12:25

7 Answers 7


Generally speaking, this genetic thief would die out really fast, or would either be so advanced as to probably not need to use its advanced processing power in this capacity.


Did I say processing power? Yes I did.

Assume we have two creatures with human-level genetic codes: on the low end this means over 20,000 genes. We are uncertain how the human genome works because to fully understand this we'd have to understand how each allele (variant of a gene) interacts with each other allele of every other gene. Already this is a hugely complex problem, but lets assume there is only a single variant of each gene (which we know not to be true, elsewise everyone would look the same, more or less), and our gene thief finds this other, human-like creature.

It has to determine how each of twenty thousand genes works with every gene in its own body. A single gene would be 20,000 computations - and we're not even looking at polygenic groupings - traits that are controlled by multiple genes. A three-gene cluster would take 20,000 * 19,999 computations. In a case where each of these three genes has two alleles, it's roughly 20,000 ^ 6, or close enough to not matter. The number of computations explodes astronomically - beyond astronomically. This gene thief has the computational prowess beyond modern reckoning. It can surely be using that to do any number of other complex things that would better ensure its survival.


The alternative is to blindly steal genetics and try them out. This is exactly the same as giving yourself cancer: cancer is just cells with different genetics replicating through your body. You're unlikely to survive: though if you did, you'd be stronger.


But! In reality genetics are swapped all the time. You absorb genetics through your food. When a virus invades your body, some of your cells learn that virus - by stealing its genetics. Babies inject their genes into their mother, causing her to become a genetic mosaic. Trees pass genes to shrubs. Genes are constantly flowing around. Most of the time, the receiving organism rejects the genes if they're dangerous. Sometimes they onboard them - and this is where gene therapy comes in. But the key is that the channels by which new genes are accepted are very narrow.

A mother pretty much knows that her child's genes are roughly the same and probably not going to kill her. Trees and plants passing trees around do die a lot, as do bacteria that swap genes - but their reproduction outpaces the death, meaning the workable imported genes end up lasting. When your immune system accepts genetics, it does so in a way explicitly designed to fight that virus. In order for the gene thief to work it would have to find a channel that narrows the odds of death by a huge amount.

And the result is unlikely to be something straightforward like, "I grow a bat's wings." Rather, it will onboard the genes and something new will emerge based on the complex interaction of the proteins that gene codes for. When gene therapy is done, they go to great lengths to determine what the result will be, and a lot of the work is around not having an adverse effect that is hard to predict. The results are not as predictable as getting whole new capabilities: instead the results will change things at a small scale (like the amount of a hormone produced), which will have different effects on different parts of the randosaur's system.

So while you can, and do, have gene thieves in the wild, it's not as 'whole feature' as you propose.


This already exists.

Bacteria already have this: they can swap small pieces of DNA called plasmids which have a few genes on them.

For example, one of the reasons that antibiotic resistance spreads through bacterial populations so fast is that the genes can be loaded on plasmids, which make it into the environment and can be picked up by other bacteria.


This will not help the creature become strong or help it survive. On the contrary, it will very probably end up killing it very soon.

The Problem Of Compatibility

You cannot steal partial genetic code from an organism and expect it to function successfully in another body. For example, if your said creature decides to live in a desert and finds it too hot to be likeable, it would naturally feel like stealing some desert resident animals' genes and make itself at home with the heat. The problem is that the genes of those animals would not kick in, doing the same thing in this creature. The problem being that genes are meaningful only in the background of the whole genetic setup of a creature. For example, it tastes some blood of a camel and decides this is a nice method for living happily in a desert (camels have a much thicker blood and lose very less water through sweating). However, stealing genes only for the camel's blood (or its skin, too and water management system) would be a bad idea because those genes function well only in the background of the complete camel genome. For example, the antibodies present in the camel's blood would not function properly with the immune system of the creature's original blood. Similiarly, camel's water management system would probably be highly incompatible with the water intake of the creature's original organs.

So all in all, no, just stealing part of genetic code of a creature and hoping it would function properly in an otherwise alien background genome will not work.

Some Cells Are Rarely Ever Replaced

For example, brain cells (some parts of it), a lot muscle cells, several types of blood cells etc. reference

This means that these cells would not be replaceable for the organism.

The Main Problem: Cancer

This is when different types of cells have different growth/mitosis rate. Considering that cells of different organisms have different growth and multiplication rate, copying anything from a quickly multiplying cells genome would spell doom for the creature as these cells would quickly form tumors and ravage the host body with a matter of few weeks.


Maybe a different approach to others, instead of stealing genetic material and integrating it into an existing pattern there is the possibility of cloning an individual's body parts and integrating them wholesale into the host body.

That gets round the trouble of cells, then it's a case of having a plastic type mind (say octopi) that can easily adapt to controlling to new appendages and having a body that does not reject foreign bodies wholesale.

So you can grow the wings of a bat, minus the body and graft it to your own.

Note: this would take a lot of energy to grow new mature body parts, so don't forget a healthy appetite to go with it!


Steal genetic material? Probably not. But control its own growth, to become like other creatures? Maybe.

Our theories of evolution say that creatures' changes over time come from mutations and selection of advantageous traits. To have controlled, significant genetic changes within a generation without "third party" devices (e.g. genetically modified viruses, radiation) seems highly unlikely -- first of all, no known creature can change its DNA selectively. The most "adaptable" organisms (e.g. cockroaches) are adaptable because they have a high rate of mutation, not because they can control their mutations to be beneficial.

Lo, enter the Tree-o-zard.

A sentient creature, shaped like a large tree. With a small version of itself for each "branch".

This creature would induce mutations in each small body/"branch" through exposing it to radiation (plausibly, it originates from some atmosphereless planet, where radiation is abundant)(definitely no shortage of radiation in space) and allowing the different mutations to consume each branch. The "trunk" (original body) would be made of a thick radiation-proof material, so that the creature's original genetic code wouldn't be damaged. Then, when a branch is found to have a good mutation, it would be assimilated into the trunk, and all other branches destroyed (as through abscission). Then, the Tree-o-zard would regrow a new "generation" of branches, and repeats the process.

"But wait!" you say. "Wouldn't this crea--"

Yes. This creature would have to be mind-bogglingly large. To get a mutation that's actually desirable would be extremely difficult. Pretend the odds are 1 in 1,000,000,000 (1 billion). Even with ten billion branches, the odds of the Tree-o-zard getting a good mutation is hardly guaranteed (I'll make an edit later when I do the math). Therefore, the process would be extremely slow, and the creature would likely be sessile.


The use of the word "stealing" imply to deprive it's original owner of something. I think that incorporating some fragment of DNA material from one organism to another do not impact in any way the donor.

I start by clarifying this idea with an example. I will then give my opinion as to why borrowing some fragment of genetic code to incorporate in an adult is unlikely to give any visible result.

An example where that word would qualify could be: If I publish a book in which I copy a few pages from a popular author, such as Shakespeare. If my book appeal more than the original, if these few pages add value to my work and if these few pages are sufficient to convince some people to not buy the original source, depriving the original author of revenue that she would otherwise get, then this is clearly stealing.

This example is purposely exaggerated. An author can be considered as dishonest even if the source of the few pages was not from a well known author. A law suit in such case would help the original author to be discovered.

To answer to the important part of the question, I think that incorporating fragment of DNA from other organism is unlikely to help in most cases. For example, if a mammal would dream to get wings like a bird, such change can not happen on an already fully grown animal. A mammals would probably prefer to use the genes from another mammals, such as the bats, instead of using those of birds. Because the closest common ancestor between birds and mammals is much farther in time (possibly more than 300 millions year), the portions of the genetic code which are similar enough to be compatible and work in a meaningful way are very small.

The specific genes needed to change, for a large scale modification such as replacing the arms and legs with a wing system similar to bats, would need to affect the growth rate of early cells, soon after they start to differentiate.

The way the cells carefully orchestrate the rate of division, migration and pre-programmed cell death to create a viable miniature copy of an adult is an incredible amount of finely tuned apparently useless portion of DNA which actually may act as timers and other part of the genetic code controlling specific chemical triggers which act as equivalent to a decision making entity when described as an algorithm.

To give an comprehensible example, suppose you would try to build a machine which reproduce the symphony of Beethoven using thousands of coo-coo clocks, each one preprogranned to hit a bell at the right time. Let's say that you could group some of these mechanical clocks in clusters which would play the part of the music that repeat. A master clock would trigger each groups, onne by one, manage to rewind the clocks that already ding one time, to allow them to be ready to replay that portion of the song later.

Basically, a large part of the genetic code which appear to be useless for an adult is needed to build this complex 3 dimensional organism made of trillions of cells, like all pluri-cellular living organism do.

Once this structure is built, it can not morph to another one. Take the example of caterpillar. The metamorphosis to become a butterfly is done by demolishing the house and building a brand new one. Every organ is dissolved and a brand new set is developed from scratch. It is like if these two phases in the lives were two distinct creature encoded in the DNA. Half of the genetic sequence knows how to develop from egg to caterpillar.

Then, all the food accumulated by the caterpillar as fat reserve is used, similar to the yellow and white of a chicken egg, to restart the growth of a single cell inside the cocoon. That cell in the dying caterpillar restart developing like a freshly fertilized egg do, starting with a few divisions, probably 5, creating 32 identical copies. Then, starting to differentiate, each cell getting an almost identical copy of all the DNA, except small parts that account for cell specialization.

The only cell that get a fully intact DNA sequence, the entire instruction to start the next generation, are the sperm and egg. Every other cell get a copy with tiny difference, some parts of the code acting as lock to prevent the neuron cell to create the same protein as the liver cell, for example. *Bad example as I learned last week that the cells in the liver have mode than one nucleus).


This already exists in a far more impressive form - enter the Tardigrade:


They are called water bears, and are almost everywhere from the great lakes to the himalayas. They can survive in extreme heat and extreme dryness and come back to life when the time is life. Tardigrades are the inspiration of Half Life 2's antagonist race - the Combine, which use host features to combine into new life forms.



You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .