How did we use AI to monitor the evolution of bacteria on earth

Roughly A trillion species Microorganisms on earth, the majority bacteria.

Bacteria consist of a single cell. They do not have bones and are not like large animals that leave clear signs in geological records, which is also possible for paleontologists to work after millions of years.

This made it very difficult for scientists to form a timeline of their early evolution. However, with the help of machine learning, we were able to fill most of the details. Our new research, Published today ScienceIt also reveals that some bacteria have improved the ability to use oxygen long before it became saturated with the world about 2.4 billion years ago.

A monumental event in the history of the world

About 4.5 billion years ago, the month occurred. Strongly. A Mars -sized object collided with the Earth and turned its surface into a melted rock. If there was life before this disaster, it was probably destroyed.

After that, the current ancestors of all living things appeared: single -celled microbes. For the first 80% of the history of life, the world took place only by these germs.

Like the evolutionary biologist Theodosius Dobzhansky, nothing in biology makes sense except for the light of evolution He said famous In 1973. But how did the evolution of life go through the early history of the world?

Comparing DNA series from the great diversity of life we ​​see today can tell us how different groups are related to each other. For Ibnstance, we people are more closely related to mushrooms and apple trees. Likewise, such comparisons can tell us how different bacterial groups are related to each other.

However, the comparison of DNA series can only take us so far. DNA comparisons do not say when evolutionary events took place in the history of the world. Once upon a time, an organism produced two puppies. One of them led to mushrooms and the other to people (and many other species). But when did this organism live? How many years ago?

One thing that geology taught us is the existence of another monumental event in the history of the world 2.4 billion years ago. At that time, the world atmosphere has changed significantly. A group of bacteria, called cyanobacteria, invented a number that will change the story of life forever: photosynthesis.

Energy harvesting from the sun gave power to its cells. However, it also produced an inappropriate waste product, oxygen gas.

For millions of years, the oxygen in the atmosphere slowly accumulated. Before this “great oxidation event”, the world had almost no oxygen, so life wasn’t ready for it. In fact, oxygen is a toxic gas for unstoppable bacteria, and therefore releasing it to the atmosphere has probably caused a mass destruction. Survival bacteria either developed to use oxygen or retreated to the recesses of the planet, which he did not penetrate.

Bacterial Tree of Life

The great oxidation event is interesting for us, not only because of its impact on the history of life, but also because it can be given an open date. We know that this was about 2.4 billion years ago – and we also know that most bacteria adapted to oxygen have to live after this event. We used this information to lay down the bacterial tree of life on dates.

We started by educating an artificial intelligence (AI) model to predict whether bacteria live with oxygen. Many bacteria that we see today use oxygen like cyanobacteria and others living in the ocean. But many are not like bacteria living in our intestines.

As the machine learning tasks continued, it was quite simple. The chemical power of oxygen changes the genome of a bacteria significantly because the metabolism of a cell is organized around the use of oxygen and therefore there are many clues in the data.

Then, in the past, we applied our machine learning models to estimate which bacteria used oxygen. This was possible, because modern techniques not only allows us to estimate how the species we see today are, but also the genes carried in the genome of each ancestor, which genes are related to how.

A surprising bending

Using the geological geological phenomenon of the great oxidation event as a “fossil” calibration point, our approach has produced a detailed timeline of bacterial evolution.

We were able to significantly improve the timing of bacterial evolution by combining the results of geology, paleontology, philogenhetics and machine education.

Our results also emerged as a surprising bending: some bacterial lineages that could use oxygen were about 900 million years before the great oxidation event. This shows that these bacteria improve the ability to use oxygen even if the atmospheric oxygen is low.

Not remarkably, our findings show that cyanobacteria actually improve the ability to use oxygen before developing photosynthesis.

This framework not only reshapes our bacterial evolutionary understanding of history, but also shows how the abilities of life develop in response to the changing environments of the world.

This article has been re -published here. Speech Creative under license. Read original article.