Matched genome-background controls refine phylogenetically supported horizontal gene transfer candidates across Actinomycetota

Bacteria are not solitary islands; they are constantly swapping genetic material with their neighbors. This process, known as horizontal gene transfer, allows a microbe to acquire new traits—such as the ability to digest a new food source or resist an antibiotic—without waiting for the slow march of evolution through reproduction. While this exchange drives the incredible diversity of the microbial world, it also creates a major headache for scientists trying to map the history of life. When a bacterium picks up a gene from a distant relative, the genetic record becomes a tangled web rather than a clear family tree. To understand how these organisms evolved, researchers must distinguish between genes that were inherited from a parent and those that were borrowed from a neighbor. The challenge is that these borrowed genes often look very similar to the native ones, making it difficult to tell if a specific piece of DNA was truly transferred recently or if the similarity is just a result of the two organisms being closely related in the first place.

A new study focusing on a large group of bacteria called Actinomycetota has uncovered a significant flaw in how scientists currently identify these genetic swaps. The researchers analyzed over a thousand bacterial genomes to find evidence of horizontal gene transfer, but they discovered that many signals previously thought to be clear cases of borrowing were actually illusions. By comparing candidate genes not just to the rest of the world, but to the rest of the organism's own genome, the team found that a gene's "foreign" appearance is often shared by thousands of other genes in the same bacterium. This means the entire genome, not just a single gene, has a strong affinity to the proposed donor group, likely due to shared ancestry or taxonomic confusion rather than a specific, localized transfer event.

The researchers began their work with a massive collection of 1,143 bacterial genomes from the Actinomycetota phylum. They used a powerful search tool to compare every protein in these genomes against a database containing nearly 5.5 million proteins from other bacteria. The goal was to find proteins that looked more like they belonged to a different genus than to their own. This initial sweep identified thousands of suspicious candidates. The team then narrowed this list down to 264 proteins that were suitable for detailed evolutionary analysis. By building family trees for these specific proteins, they found that 77 of them formed strong, well-supported groups with a proposed donor lineage. At this stage, the evidence looked promising: these proteins had clear evolutionary links to a different group of bacteria, suggesting a successful transfer.

However, the study took a crucial next step that changed the interpretation of the results. The researchers realized that finding a gene that looks like it came from a donor is not enough; they needed to know if that gene was unique in its foreignness. To test this, they looked at the immediate neighborhood of each of the 77 candidate genes. In many cases, the genes surrounding the candidate also looked like they belonged to the donor group, creating a pattern that suggested a large block of DNA had been transferred together. But the team went further. They sampled thousands of other proteins from the same bacterial genomes, far away from the candidate genes, to see if these random proteins also showed a preference for the donor group.

The results of this comparison were striking. When the researchers checked the background proteins, they found that in 35 of the 77 cases, the entire genome showed a strong affinity to the proposed donor. In other words, the "foreign" signal was not a localized event; it was a property of the whole organism. This suggested that the initial signal was likely caused by the two groups of bacteria being closely related or by gaps in the scientific database, rather than a specific gene transfer. After filtering out these genome-wide effects, only 13 proteins remained that were truly exceptional. These 13 proteins were locally enriched, meaning they were significantly more similar to the donor group than the rest of the genes in their own genome. These 13 proteins collapsed into 10 distinct candidate events, representing the most reliable signs of horizontal gene transfer found in the study.

Among these final candidates, two stood out as particularly strong examples. One involved a transfer between Calidifontibacter and Naumannella, where two proteins in the same location showed an overwhelming preference for the donor lineage, with no signs of assembly errors or mobile genetic elements that could confuse the results. Another strong signal appeared between Gordonibacter and Eggerthella. The study also highlighted several cases that required caution. Some candidates were located on short, fragmented pieces of DNA or near mobile genetic elements like plasmids, which can make it difficult to confirm the transfer. Others involved bacteria whose names have recently changed due to new scientific understanding, meaning the "donor" and "recipient" might be more closely related than their current names suggest.

The study concludes that relying solely on the similarity of a gene to a foreign group can lead to false conclusions, especially in groups of bacteria that are closely related and taxonomically unstable. The researchers argue that the best way to confirm a gene transfer is to use the recipient's own genome as a control. If a gene looks foreign, but the rest of the genome looks just as foreign, then no specific transfer has been proven. By applying this stricter standard, the team refined a list of thousands of potential candidates down to a small, high-priority set of ten events. These ten events represent the most credible examples of horizontal gene transfer in this group of bacteria, offering a clearer picture of how these organisms have exchanged genetic material to adapt and evolve. The findings serve as a reminder that in the complex world of bacterial evolution, the context of the whole genome is just as important as the signal of a single gene.