Viruses are in all places. They’re within the air; in sewage, lakes, and oceans; in grasslands and decaying wood. Some thrive in extreme conditions, like hydrothermal vents, Antarctic ice, and potentially even outer space.
They’re also ancient. Some are likely as old as, if not even older than, the very first cells.
Despite cohabitating with viruses because the dawn of our species, the viral universe stays largely mysterious. For many years, scientists have painstakingly gathered samples from across the globe and sequenced their genetic material. But viruses rapidly mutate, and these efforts only scrape the surface of the virosphere.
Most viral genetic material is biological “dark matter,” Mang Shi at Sun Yat-sen University and colleagues recently wrote in a brand new paper published in Cell.
With the assistance of AI, the team is shedding recent light on the viral world. The AI, dubbed LucaProt, relies on a big language model to make sense of chunks of viral genetic material. One other algorithm further parses genetic data into more “digestible” bits to extend efficacy.
After analyzing nearly 10,500 samples—some from previous databases, others collected through the study—the AI detected 70,458 recent RNA viruses from samples everywhere in the globe.
“Swiftly you possibly can see things that you just just weren’t seeing before,” Artem Babaian on the University of Toronto, who wasn’t involved within the study, told Nature.
Viruses have a nasty repute. The Covid-19 pandemic and annual flu season highlight their destructive side. But they will also be used to battle antibiotic-resistant bacteria, shuttle gene therapies into cells, or be developed into vaccines.
Charting the viral universe offers a bird’s-eye view on the evolution and mutation of viruses—with implications not only for biotechnology but potentially for battling the following pandemic too.
Going Viral
In humans, DNA carries the genetic blueprint. DNA translates to RNA—also made up of 4 genetic letters—which carries the genetic information right into a cellular factory to make proteins.
Viruses are different. Some forgo DNA altogether, as an alternative directly encoding their genetic blueprint in RNA. It sounds unusual, but you already know a few of these viruses: SARS-CoV-2, which causes Covid-19, is an RNA virus. These viruses have proteins that science knows little about, and so they could also offer recent insight into biology.
For many years, scientists have tried to decode the virosphere by collecting samples. The sources range from the on a regular basis—water from an area creek—to the acute, akin to Antarctic ice or deep seawater. RNA extracted from these samples is rigorously sequenced and deposited into databases. This method, called metagenomics, captures snippets of all viral RNA from an environment.
Making sense of the genetic goldmine takes more work. Classic computational methods struggle to sift these large databases for meaningful insights.
Enter ESMFold. Developed by Meta, this system relies on large language models—the identical technology powering OpenAI’s ChatGPT and Google’s Gemini—to predict protein structures based on their amino acid “letters.” Similar methods, including DeepMind’s AlphaFold and David Baker’s RoseTTAFold, recently won their developers the 2024 Nobel Prize in Chemistry.
ESMFold takes in molecular sequences and predicts the 3D structures of proteins on the atomic level. For its first real-life task, scientists used the AI to decode the “dark matter” of proteins in microbes we all know the least about. Last yr, the AI predicted the structure of over 700 million proteins from microorganisms. Ten percent were completely alien to any previously discovered.
Taking note, Shi’s team asked if an identical strategy could work on the earth of RNA viruses.
Panning for Viruses
Scientists have previously used AI to fish out potential recent RNA viruses from petabytes of genetic sequencing data—an amount roughly such as 500 million high-resolution photos.
These studies focused on RNA-dependent RNA polymerase, or RdRP. Here, the RNA sequences encode RdRPs, a family of proteins that tags most RNA virus genomes. An early evaluation identified nearly 132,000 recent RNA viruses based on their genetic data.
The issue? Viruses rapidly mutate. If the genetic letters encoding RdRPs change, AI trained on those sequences may not find a way to acknowledge mutated viruses. The brand new study tackled the issue by marrying the previous approach with ESMFold in a two-channel AI.
The primary channel uses a transformer-based model, just like ChatGPT, to extract amino acid sequence “keywords” encoding viral RdRPs from a big database. After training with the specified sequences, and a few that were randomly generated, the AI created a vocabulary of about 20,000 continuously occurring protein sequences encoding for RdRPs.
In comparison with previous methods, this step breaks genetic libraries into more digestible sections, making it easier for the AI to tackle longer genetic sequences and detect viral RdRP proteins.
The second channel taps a version of ESMFold. That is the slow but careful reader. Reasonably than blazing through protein words, it “reads” each letter and predicts how each structurally connects with others to form 3D protein shapes. This step grounds the AI, giving it an idea of how RdRPs should look in living viruses.
LucaProt was trained on nearly 6,000 sequences encoding RdRP proteins and over 229,500 sequences known to encode different proteins. Challenged with a test dataset, wherein the researchers knew the answers, the AI was exceptionally accurate, returning false positives only 0.014 percent of the time.
The AI found 70,458 potential recent, unique viruses. One, isolated from dirt, had a surprisingly long genome—”one in all the longest RNA viruses identified thus far,” wrote the team. Others could thrive in hot springs and very salty lakes.
The expanded virosphere adds recent viruses to known viral groups—for instance, Flaviviridae, which causes hepatitis or yellow fever. LucaProt also identified 60 different viral groups, each highly different than all known viruses today.
It’s to not say they cause diseases, but they “have largely been ignored in previous RNA virus discovery projects,” wrote the team.
To Babaian, the study found “little pockets of RNA virus biodiversity which are really far off within the boonies of evolutionary space.”
A Viral Hit?
Viruses require a living host to survive. The team is upgrading their AI to predict these hosts. Most RNA viruses infect eukaryotes, which include plants, animals, and humans. Some viruses may infect bacteria—their cat-and-mouse game inspired the gene editor CRISPR-Cas9.
“The evolutionary history of RNA viruses is at the least as long, if not longer, than that of the cellular organisms,” wrote the authors.
Often ignored is the third branch of life, archaea. Evolved through the early stages of life on Earth, these lifeforms share similarities to bacteria and eukaryotes—for instance, how their genetic material replicates.
But archaea are a definite branch of life that thrives in extreme environments, akin to hydrothermal vents or extremely salty water. There are hints that RNA viruses could also infect archaea. In that case, it could spur recent insights into our tree of life—and as with CRISPR, potentially result in recent biotechnologies.
