1 Trillion Species, 3 Billion Years: How to Track the Evolution of Bacteria on Earth using AI

There are about 1 trillion species of microorganisms on Earth. Most of it is bacteria.

Bacteria are made up of a single cell. They have no bones and are not like large animals that leave clear signs in the geological records. This is something paleontologists who appreciate it in a few million years can study.

This made it extremely difficult for scientists to establish a timeline for early evolution. But with the help of machine learning, we were able to fill in many details. Our new research published today in Science reveals that it has also developed several bacteria that developed the ability to use oxygen long before the planet saturated about 2.4 billion years ago.

Monumental events in the history of the earth

About 4.5 billion years ago, the moon formed. Hard. A Mars-sized object collided with Earth, turning its surface into molten rock. If life had existed before this cataclysm, it would have probably been destroyed.

The present ancestors of all organisms then appeared: single-cell microorganisms. In the first 80% of life history, only these microorganisms lived on Earth.

As evolutionary biologist Theodosius Dob Hansky famously said in 1973, there is no meaning in biology except for the light of evolution.

Comparing DNA sequences with the incredible diversity of diversity we see today will teach you how different groups are related to each other. In the case of ibnstance, we humans are more closely associated with mushrooms than with apple trees. Similarly, such comparisons tell us how different groups of bacteria are related to each other.

However, comparisons of DNA sequences can take us so far. DNA comparisons do not show when the evolutionary events of Earth’s history took place. At one point, the organism recreated two offspring. One of them produced mushrooms, and the other occurred in humans (and many other species too). But when did the creature live? How many years ago?

One thing geology teaches us is the existence of another monumental event in the history of the Earth 2.4 billion years ago. At that time, the Earth’s atmosphere changed dramatically. A group of bacteria called cyanobacteria invented a trick that will change the story of life forever: photosynthesis.

When energy was harvested from the sun, the cells were powered. However, it also produced inconvenient waste, oxygen gas.

For millions of years, oxygen in the atmosphere slowly accumulated. Before this “great oxidation event,” life was unprepared as the Earth contained little oxygen. In fact, for beginner bacteria, oxygen is a toxic gas, so release into the atmosphere probably caused a massive extinction. Surviving bacteria either evolved to use oxygen or retreated into planetary depressions that do not penetrate.

Bacteria Tree of Life

Large oxidation events are of particular interest to us as they can not only impact life history, but also provide clear dates. We know it happened about 2.4 billion years ago. We also know that most bacteria that were adapted to oxygen had to survive after this event. Using this information, I dated the bacteria tree of life.

We started by training an artificial intelligence (AI) model to predict whether bacteria are alive with oxygen. Many of the bacteria we see today use oxygen, such as cyanobacteria. But many of the bacteria that live in our gut.

As far as machine learning tasks go, this was pretty easy. There are many cues in the data as cellular metabolism is organized around the use of oxygen, and the chemical ability of oxygen significantly alters the bacterial genome.

Next, we applied a machine learning model to predict which bacteria were used in the past. This was possible because modern technology allowed us to infer not only how the species we see today are related, but also the genes that each ancestor carried in its genome.

An amazing twist

By effectively using planetary-wide geological events with large oxidation events as a “fossil” calibration point, our approach has produced a detailed timeline of bacterial evolution.

Combining geology, paleontology, systematics and machine learning results has been able to significantly improve the timing of bacterial evolution.

Our results also revealed an incredible twist. Several bacterial strains that can use oxygen were present around 900 million years before the major oxidation event. This suggests that these bacteria have evolved their ability to use oxygen even when they are deficient in atmospheric oxygen.

Surprisingly, our findings showed that cyanobacteria evolved their ability to use oxygen before they actually developed photosynthesis.

This framework not only understands the history of bacterial evolution, but also shows how the capabilities of life evolved in response to the changing environment of the planet.

This article will be republished from the conversation under a Creative Commons license. Please read the original article.