Self-replication

Some things can make copies of themselves.

Small parts of your body do this. They make more parts to help you grow.

Living cells do this by dividing. They copy their tiny plans to pass to new life. Viruses also make copies. They do this by using a cell.

Computer programs can copy themselves too. Even some shapes can do this. You can join small shapes to make a bigger one.

Sometimes the copies are not perfect. These small changes can help things live better. This helps life change over time.

It is amazing how things copy themselves!

Self-replication is when a system makes a copy of itself.

Living cells do this through cell division. During this way, they copy their DNA. DNA is the code that holds instructions for life. This code can pass to new life. Viruses also replicate. They do this by taking over the parts of a cell.

Sometimes, the copies are not perfect. We call these changes mutations. These changes can make a copy different from the first one. Some changes help a living thing survive better in its home. This leads to natural selection. This is how life changes over time.

Even shapes and computers can copy themselves. In math, some shapes are called rep-tiles. You can join small tiles to make a larger version of the same shape.

Scientists also study how robots might copy themselves. A robot would need to find parts and make new ones. This could help make things like tools much cheaper. Some people even study how clay crystals grow and copy tiny errors.

Self-replication is a special way that a system makes a copy of itself.

This happens in many different ways in our world. Biological cells use cell division to create new versions of themselves. During this time, they also copy their DNA. DNA is the set of instructions that tells a living thing how to grow. Even viruses use this idea to replicate. They do this by taking over the machinery inside a cell. Some tiny proteins called prions can even replicate by changing normal proteins into new forms.

How does a system actually build a copy? Early research by a scientist named John von Neumann helped explain this. He found that a replicator needs a few specific parts. First, it needs a coded version of itself. Next, it needs a way to copy that code. Finally, it needs a way to build the new copy using things in its environment.

Scientists have even tried to make this happen with DNA in a lab. In 2021, researchers used sixteen different DNA sequences to make a system that copies itself. They changed the temperature up and down to make the copies grow.

History shows us that self-replication is a very old part of life. Some thinkers believe it started with a molecule similar to RNA. This molecule could split into two pieces to act as a template for new strands. This process helped create the first simple living things. Because these early copies were not always perfect, they had mutations. A mutation is a tiny change in the code. These changes meant some versions were better at surviving than others. This led to natural selection, which is how life changes over time.

There are many different classes of things that can replicate. Natural replicators are living things that use designs from nature. Some scientists think humans could design autotrophic replicators. These would be able to find their own materials in the wild. There are also ideas for machines that use metal bars and wires to build copies.

In math, we see self-replication in shapes called rep-tiles. One example is the sphinx hexiamond. You can join four of these shapes together to make a larger version of the same shape.

Today, people study self-replication to help with many hard jobs. Engineers want to build "clanking replicators," which are machines that copy themselves. This could make making tools much cheaper. NASA is even looking at ways for machines to mine resources in space. In computer science, a program called a quine can print its own code.

Whether it is a tiny cell or a large robot, the idea of making a copy is a powerful way to grow and change.

Self-replication is a fundamental behavior of a dynamical system. It occurs when a system produces an identical or similar copy of itself. This process is essential for the continuation of life and the growth of complex systems. In biology, cells reproduce through cell division. During this process, DNA is replicated so it can be passed to offspring.

Even non-living things can show these traits. Viruses replicate by hijacking the reproductive machinery of a host cell. Harmful prion proteins replicate by converting normal proteins into rogue forms. In the digital world, computer viruses use existing hardware and software to reproduce.

To understand how this works, we can look at the theory of John von Neumann. His research established that replicators generally require three specific components. First, they need a coded representation of the replicator itself. Second, they need a mechanism to copy that coded information. Third, they need a mechanism to carry out construction within the host environment.

Scientists have attempted to recreate this in laboratories. In 2021, researchers built a system using sixteen specially designed DNA sequences. By changing the temperature up and down, they could link four sequences together following a template. This increased the number of template copies in each cycle without needing an external enzyme.

Self-replication is considered a core feature of the origin of life. One theory suggests it emerged when a molecule similar to a double-stranded polynucleotide, perhaps RNA, dissociated into single strands. Each strand then acted as a template to synthesize a complementary strand. This produced two double-stranded copies from the original.

These early replicators could compete for available mononucleotide resources. If a sequence was more "fit," it would survive better. Because these early processes were often inaccurate, they produced mutations. These mutations caused genetic variation, allowing natural selection to drive the evolution of life.

Researchers categorize replicators based on how much support they require. Natural replicators, like living organisms, use designs from nonhuman sources. Autotrophic replicators are systems that can reproduce "in the wild" by mining their own materials. Scientists conjecture that humans could design non-biological autotrophic replicators to create specific products. Self-reproductive systems might use industrial feedstocks like metal wire to build copies. Finally, self-assembling systems use finished, delivered parts to build themselves. A study by Robert Freitas and Ralph Merkle identified 137 design dimensions for machine replicators, including replication control and energetics.

Self-replication also appears in the field of geometry through self-replicating tilings. These are patterns where several congruent tiles join to form a larger tile similar to the original. This is part of a study called tessellation. One famous example is the "sphinx" hexiamond, which is the only known self-replicating pentagon.

By joining four of these shapes, you can create a larger sphinx. Solomon W. Golomb called these "rep-tiles." Lee Sallows later identified them as a special type of "setiset." A setiset of order $n$ uses $n$ shapes to form larger replicas. A "perfect" setiset is one where every shape in the set is distinct.

In computer science, a self-reproducing program is called a quine. When executed, a quine outputs its own source code. A simpler method involves a program that copies any data stream directed at it. If that stream is the program's own code, it becomes self-reproducing. This is similar to how biological life works, where the program is treated as both executable code and data. This approach is often simpler because the program does not need to contain a complete description of itself.

Engineers are interested in creating "clanking replicators," which are mechanical devices that can copy themselves. The goal is to lower the cost of manufactured goods by avoiding labor and distribution costs. A NASA study estimated that the complexity of such a device is similar to an Intel Pentium 4 CPU. This suggests the technology is achievable with a small engineering group. In robotics, a self-replicating machine would need to obtain materials, manufacture parts, provide power, and correct errors. This research connects to many fields, from nanotechnology to space resource mining, where NASA studies machines that can copy themselves to work in space.