Opening details

A new study suggests that interactions between genes play a larger role than expected in determining how a genome changes. Some regions of the genome are known to be more prone to mutations than others, but the evolutionary history of a species may also play a role in increasing the predictability of mutations.

The results of this research are revolutionary. By showing that evolution is not as random as we previously thought, we have opened the door to rich possibilities in synthetic biology, medicine, and environmental science.
– says evolutionary biologist James McInerney from the University of Nottingham.

Biologist Alan Beavan from the University of Nottingham and his colleagues used the computational capabilities of artificial intelligence to examine more than 2,000 complete genomes of Escherichia coli bacteria, which live in the guts of warm-blooded organisms.

Bacteria are particularly crafty at changing their DNA; They are very adept at stealing genes from the environment and incorporating them into their own genomes. This process, known as horizontal gene transfer, allows bacteria to freely access new traits, such as antibiotic resistance, without having to wait long for selection to act over several generations.

Interestingly, horizontally transferred genes belonging to the same basic group can be found in different places in the bacterial genome. By examining horizontal genes in different positions, the researchers were able to see how they were affected by the genes’ immediate environment. They were able to test the imaginary experiment of famous evolutionary biologist Stephen Gould: Replaying the tape of evolutionary history over and over again will lead to different, unpredictable results each time because evolutionary paths depend on unpredictable events.

If this were the case, the bacterial genome would continue to evolve chaotically after the acquisition of a new horizontal gene. But after these gene acquisition events, AI discovered patterns of predictability in these thousands of “tape repeats.” “We found that some gene families do not appear in the genome at all when another particular gene family is already there, and in other cases some genes are highly dependent on the presence of another gene family.”– explains researcher Maria Rosa Domingo-Sananes from the University of Nottingham.

So the history of the genome, that is, the number of genes it has now, can determine which genes it will or will not have in the future. Scientists have seen hints of this before, when genes located physically close to each other on genetic molecules (linked genes) are lost or gained together. But this also happened in genes that did not have a close physical connection in bacterial genomes.

Some aspects of evolution are deterministic; In other words, these are likely to happen every time we listen to the tape. The presence or absence of a gene can only be predicted based on other genes in the genome. For example, a hypothetical gene A can predict the presence of gene B only in the absence of gene C.
Beavan and his team confirm this in their paper.

This does not violate the random mutation rule; Rather, it means that the forces of natural selection are also at work at the molecular level, and until recently we did not have the computational power to see this fully. In fact, genomes are their own microscopic ecosystems in which genes can help or hinder each other.

Based on this work, it is possible to begin to investigate how genes support resistance to antibiotics, for example, and in the future produce drugs against diseases that cannot be treated today.