In a groundbreaking study published in the Proceedings of the National Academy of Sciences (PNAS), scientists at Pacific Northwest Research Institute (PNRI) have overturned a long-held belief in genetics: that inheriting two harmful variants in the same gene always worsens disease. Instead, the team found that, in many cases, two harmful variants can actually restore normal protein function.
The research focused on a human enzyme called argininosuccinate lyase (ASL), which plays a critical role in removing toxic ammonia from the body. Variants in ASL that decrease its activity cause one of the urea cycle disorders, a set of rare and potentially life-threatening metabolic diseases.
By experimentally measuring the functional impact of several thousand individual variants and variant combinations, PNRI researchers discovered that over 60% of pairs that were individually damaging could, together, bring enzyme activity back to healthy levels.
“This work shows that genetic variants don’t act independently in many important cases,” said Michelle Tang, PhD, PNRI Staff Scientist and lead author of the study. “For a defined group of genes, the default assumptions we use to predict disease risk simply don’t hold.”
This phenomenon is known as intragenic complementation. It occurs when damage caused by one variant is offset by another variant in a different part of the same protein. The mechanism was first proposed in 1964 by Francis Crick and Leslie Orgel, but until now had not been tested systematically or shown to be common or predictable at scale.
To make these interactions predictable, the research team developed an AI-based model that accurately forecasted whether two variants would restore protein function. The model achieved nearly 100% accuracy in predicting intragenic complementation in ASL as well as in a second human enzyme, fumarase, suggesting these rules apply broadly across the human genome.
“We’ve shown that, in many cases, two damaging variants can work together to restore protein function,“ said Aimée Dudley, PhD, PNRI Senior Investigator who led the study. “This kind of genetic interaction is not an isolated exception, but a widespread and underappreciated way that variants can interact, especially in rare disease contexts.”
The researchers estimate that approximately 4% of human genes have the structural features that allow this type of interaction. For these genes, standard genetic predictions can overestimate disease risk, particularly for people who carry two different variants in the same gene.
Cedars-Sinai scientists have developed an experimental drug that repairs DNA and serves as a prototype for a new class of medications that fix tissue damage caused by heart attack, inflammatory disease or other conditions.
“By probing the mechanisms of stem cell therapy, we discovered a way to heal the body without using stem cells,” said Eduardo Marbán, MD, PhD, executive director of the Smidt Heart Institute at Cedars-Sinai and the study’s senior author. “TY1 is the first exomer – a new class of drugs that address tissue damage in unexpected ways.”
TY1 is a laboratory-made version of an RNA molecule that naturally exists in the body. The research team was able to show that TY1 enhances the action of a gene called TREX1, which helps immune cells clear damaged DNA. In so doing, TY1 repairs damaged tissue.
The development of TY1 has been more than two decades in the making. It started when Marbán’s previous laboratory at Johns Hopkins University developed a technique to isolate progenitor cells from the human heart. Like stem cells, progenitor cells can turn into new healthy tissue, but in a more focused manner than stem cells. Heart progenitor cells promote the regeneration of the heart, for example.
Later, at Marbán’s lab at Cedars-Sinai, Ahmed Ibrahim, PhD, MPH, discovered that these heart progenitor cells send out tiny molecule-filled sacs called exosomes. These sacs are loaded with RNA molecules that help repair and regenerate injured tissue.
Ahmed Ibrahim, PhD, MPH“Exosomes are like envelopes with important information,” said Ibrahim, who is associate professor in the Department of Cardiology in the Smidt Heart Institute and first author of the paper. “We wanted to take apart these coded messages and figure out which molecules were, themselves, therapeutic.”
Scientists genetically sequenced the RNA material inside the exosomes. They found that one RNA molecule was more abundant than the others, hinting it might be involved in tissue healing. The investigators found the natural RNA molecule to be effective in promoting healing after heart attacks in laboratory animals. TY1 is the synthetic, engineered version of that RNA molecule, designed to mimic the structure of approved RNA drugs already in the clinic. TY1 works by increasing the production of immune cells that reverse DNA damage, a process that minimises the formation of scar tissue after a heart attack.
“By enhancing DNA repair, we can heal tissue damage that occurs during a heart attack,” Ibrahim said. “We are particularly excited because TY1 also works in other conditions, including autoimmune diseases that cause the body to mistakenly attack healthy tissue. This is an entirely new mechanism for tissue healing, opening up new options for a variety of disorders.”
The investigators next plan to study TY1 in clinical trials.
Researchers at the University of Tokyo and other institutions have made a surprising discovery hiding in people’s mouths: Inocles, giant DNA elements that had previously escaped detection. These appear to play a central role in helping bacteria adapt to the constantly changing environment of the mouth. The findings provide fresh insight into how oral bacteria colonise and persist in humans, with potential implications for health, disease and microbiome research.
You might think that modern medical science knows everything there is to know about the human body. But even within the last decade, small, previously unknown organs have been discovered, and there’s one area of human biology that is currently going through a research renaissance, the microbiome. This includes familiar areas such as the gut microbiome, but also the oral microbiome. Inspired in part by recent discoveries of extraneous DNA in the microbiome of soil, Project Research Associate Yuya Kiguchi and his team turned their sights to a large set of saliva samples collected by the Yutaka Suzuki Lab of the Graduate School of Frontier Sciences at the University of Tokyo. They wondered if they might find something similar in human saliva.
“We know there are a lot of different kinds of bacteria in the oral microbiome, but many of their functions and means of carrying out those functions are still unknown,” said Kiguchi. “By exploring this, we discovered Inocles, an example of extrachromosomal DNA – chunks of DNA that exist in cells, in this case bacteria, but outside their main DNA. It’s like finding a book with extra footnotes stapled to it, and we’re just starting to read them to find out what they do.”
Detecting Inocles was not easy, as conventional sequencing methods fragment genetic data, making it impossible to reconstruct large elements. To overcome this, the team applied advanced long-read sequencing techniques, which can capture much longer stretches of DNA. A key breakthrough came from co-first author Nagisa Hamamoto, who developed a method called preNuc to selectively remove human DNA from saliva samples, greatly improving the quality of sequencing long sections of other DNA. This allowed the researchers to assemble for the first time complete Inocle genomes, which turned out were hosted by the bacteria Streptococcus salivarius, though identifying the host itself was a difficult matter.
“The average genome size of Inocle is 350 kilobase pairs, a measure of length for genetic sequences, so it is one of the largest extrachromosomal genetic elements in the human microbiome. Plasmids, other forms of extrachromosomal DNA, are at most a few tens of kilobase pairs,” said Kiguchi. “This long length endows Inocles with genes for various functions, including resistance to oxidative stress, DNA damage repair and cell wall-related genes, possibly involved in adapting to extracellular stress response.”
The team aims to develop stable methods for culturing Inocle-containing bacteria. This will allow them to investigate how Inocles function, whether they can spread between individuals, and how they might influence oral health conditions such as cavities and gum disease. Since many Inocle genes remain uncharacterised, researchers will use a mixture of laboratory experiments and also computational simulations such as AlphaFold to predict and model the roles Inocles may play.
“What’s remarkable is that, given the range of the human population the saliva samples represent, we think 74% of all human beings may possess Inocles. And even though the oral microbiome has long been studied, Inocles remained hidden all this time because of technological limitations,” said Kiguchi. “Now that we know they exist, we can begin to explore how they shape the relationship between humans, their resident microbes and our oral health. And there’s even some hints that Inocles might serve as markers for serious diseases like cancer.”
University of Hawaiʻi at Mānoa scientists have uncovered a direct link between a missing Y chromosome gene and male infertility. Their new research reveals that deleting this single gene in mice not only caused infertility but also disrupted hundreds of other genes vital for healthy sperm. The findings, published August 27 in Cell Death and Differentiation, offer significant implications for understanding reproductive health.
Using CRISPR gene-editing, the team created mice missing one or both versions. Males without both, known as Zfy double knockouts, were completely infertile, with severely abnormal or absent sperm.
“This work really pushes forward our understanding of how this important Zfy gene works,” said Ward. “We identified pathways and other genes that are affected and we can now study how exactly Zfy regulates them.”
To continue investigations, the researchers turned to assisted reproduction techniques pioneered at UH, including intracytoplasmic sperm injection (ICSI) and round spermatid injection (ROSI). This allowed them to examine the molecular consequences of Zfy loss.
When one gene disrupts hundreds
The results revealed that without Zfy, hundreds of genes became misregulated – some too active, others too weak. Many of these genes are responsible for sperm production, DNA packaging, and cell survival.
As a result, sperm precursor cells in the testes died off early, and the sperm that did form carried fragile DNA that wasn’t properly condensed.
The study details can be found in an article published in Cell Death and Differentiation, a leading peer-reviewed journal.
Couple or siblings? New study may explain why we prefer partners who are similar to us. Photo by Daniil Onischenko on Unsplash
It is no secret that people are often drawn to romantic partners who seem similar to themselves. This tendency, called assortative mating, has been established in humans (Horwitz et al., 2023; Luo, 2017) as well as other species. Fish, for example, demonstrate the behaviour frequently (Jiang et al., 2013).
Assortative mating has also recently been in focus on social media with the viral Siblings or Dating game, where people guess whether two individuals who look alike are related or a couple.
The idea is well-founded in academic research. Humans have been observed to select partners with similar physical, personality, and demographic traits (Horwitz et al., 2023), which can impact the genetics of populations – creating subgroups that emphasise the presence of shared traits (Abdellaoui et al., 2015).
But selecting a partner like ourselves may not be solely determined by personal choice. A new study soon to be published in Psychological Science suggests that assortative mating can be explained relatively simply by looking at the inheritance of preferred traits and corresponding preferences for those traits.
Coauthors Kaitlyn Harper and Brendan Zietsch from the University of Queensland describe this scenario simply: If you are tall, you may have inherited tallness from one parent (say, your mother) and the preference for tallness in a romantic partner from your other parent (in this case, your father). The combination of those inherited traits means that you exist in the world as a tall person and are attracted to tall people.
The idea that preference for a particular trait could lead to genetic correlations has been discussed in previous research but is a newer concept for evolutionary psychology, especially in the context of assortative mating.
“The pieces were there, but they hadn’t been connected in this way before,” Harper said. “Agent-based modelling helped us connect the dots – by simulating populations, we could see that assortative mating naturally emerged without the need for additional assumptions or processes.”
She added that this research wouldn’t have been possible without an interdisciplinary mindset.
“The mechanism itself is familiar in evolutionary biology, but it wasn’t thought of as an explanation for assortative mating,” she said. “Making that connection only became possible when we looked across the two disciplines.”
To test this theory, the authors ran an agent-based model where partners are chosen according to heritable traits and preferences over 100 generations. They included models with and without selection pressure on the number of offspring within each generation to assess how the theory stands up under more naturalistic conditions.
They found that even with up to 10 preferences for traits in a partner, clear genetic correlations formed between traits and preferences for those traits, which resulted in the agents choosing partners similar to themselves. Models with selection pressure generated less-stable correlations, which the authors attribute to reduced variance in traits.
“The power of this finding is in its parsimony – it shows that a phenomenon which has puzzled researchers for decades can be understood through an explanation that was hiding in plain sight,” Harper said. “And because the mechanism is so general, it can also apply to assortative mating in animals, where many of the explanations proposed for humans wouldn’t make sense.”
Animals that hibernate are incredibly resilient. They can spend months without food or water, muscles refusing to atrophy, body temperature dropping to near freezing as their metabolism and brain activity slow to a crawl. When they emerge from hibernation, they recover from dangerous health changes similar to those seen in type 2 diabetes, Alzheimer’s disease, and stroke.
New genetic research suggests that hibernating animals’ superpowers could lie hidden in human DNA – with clues on how to unlock them, perhaps one day leading to treatments that could reverse neurodegeneration and diabetes.
A gene cluster called the “fat mass and obesity (FTO) locus” plays an important role in hibernators’ abilities, the researchers found. Intriguingly, humans have these genes too. “What’s striking about this region is that it is the strongest genetic risk factor for human obesity,” says Chris Gregg, PhD, professor in neurobiology and human genetics at University of Utah Health and senior author on the studies. But hibernators seem able to use genes in the FTO locus in new ways to their advantage.
The team identified hibernator-specific DNA regions that are near the FTO locus and that regulate the activity of neighbouring genes, tuning them up or down. The researchers speculate that adjusting the activity of neighbouring genes, including those in or near the FTO locus, allows hibernators to pack on the pounds before settling in for the winter, then slowly use their fat reserves for energy throughout hibernation.
Indeed, the hibernator-specific regulatory regions outside of the FTO locus seem crucial for tweaking metabolism. When the researchers mutated those hibernator-specific regions in mice, they saw changes in the mice’s weight and metabolism. Some mutations sped up or slowed down weight gain under specific dietary conditions; others affected the ability to recover body temperature after a hibernation-like state or tuned overall metabolic rate up or down.
Intriguingly, the hibernator-specific DNA regions the researchers identified weren’t genes themselves. Instead, the regions were DNA sequences that contact nearby genes and turn their expression up or down, like an orchestra conductor fine-tuning the volume of many musicians. This means that mutating a single hibernator-specific region has wide-ranging effects extending far beyond the FTO locus, explains Susan Steinwand, research scientist in neurobiology at U of U Health and first author on one of the studies. “When you knock out one of these elements – this one tiny, seemingly insignificant DNA region – the activity of hundreds of genes changes,” she says. “It’s pretty amazing.”
Understanding hibernators’ metabolic flexibility could lead to better treatments for human metabolic disorders like type 2 diabetes, the researchers say. “If we could regulate our genes a bit more like hibernators, maybe we could overcome type 2 diabetes the same way that a hibernator returns from hibernation back to a normal metabolic state,” says Elliott Ferris, MS, bioinformatician at U of U Health and first author on the other study.
Uncovering the regulation of hibernation
Finding the genetic regions that may enable hibernation is a problem akin to excavating needles from a massive DNA haystack. To narrow down the regions involved, the researchers used multiple independent whole-genome technologies to ask which regions might be relevant for hibernation. Then, they started looking for overlap between the results from each technique.
First, they looked for sequences of DNA that most mammals share but that had recently changed in hibernators. “If a region doesn’t change much from species to species for over 100 million years but then changes rapidly and dramatically in two hibernating mammals, then we think it points us to something that is important for hibernation, specifically,” Ferris says.
To understand the biological processes that underlie hibernation, the researchers tested for and identified genes that turn up or down during fasting in mice, which triggers metabolic changes similar to hibernation. Next, they found the genes that act as central coordinators, or “hubs,” of these fasting-induced changes to gene activity.
Many of the DNA regions that had recently changed in hibernators also appeared to interact with these central coordinating hub genes. Because of this, the researchers expect that the evolution of hibernation requires specific changes to the controls of the hub genes. These controls comprise a shortlist of DNA elements that are avenues for future investigation.
Awakening human potential
Most of the hibernator-associated changes in the genome appeared to “break” the function of specific pieces of DNA, rather than confer a new function. This hints that hibernators may have lost constraints that would otherwise prevent extreme flexibility in the ability to control metabolism. In other words, it’s possible that the human “thermostat” is locked to a narrow range of continuous energy consumption. For hibernators, that lock may be gone.
Hibernators can reverse neurodegeneration, avoid muscle atrophy, stay healthy despite massive weight fluctuations, and show improved aging and longevity. The researchers think their findings show that humans may already have the needed genetic code to have similar hibernator-like superpowers—if we can bypass some of our metabolic switches.
“Humans already have the genetic framework,” Steinwand says. “We just need to identify the control switches for these hibernator traits.” By learning how, researchers could help confer similar resilience to humans.
Genetic analysis of the early pandemic virus shows key adaptations to humans.
Creative artwork featuring colourised 3D prints of influenza virus (surface glycoprotein hemagglutinin is blue and neuraminidase is orange; the viral membrane is a darker orange). Note: Not to scale. Credit: NIAID
Researchers from the universities of Basel and Zurich have used a historical specimen from UZH’s Medical Collection to decode the genome of the virus responsible for the 1918-1920 influenza pandemic in Switzerland. The genetic material of the virus reveals that it had already developed key adaptations to humans at the outset of what became the deadliest influenza pandemic in history.
New viral epidemics pose a major challenge to public health and society. Understanding how viruses evolve and learning from past pandemics are crucial for developing targeted countermeasures. The so-called Spanish flu of 1918-1920 was one of the most devastating pandemics in history, claiming some 20 to 100 million lives worldwide. And yet, until now, little has been known about how that influenza virus mutated and adapted over the course of the pandemic.
More than 100-year-old flu virus sequenced
An international research team led by Verena Schünemann, a paleogeneticist and professor of archaeological science at the University of Basel (formerly at the University of Zurich) has now reconstructed the first Swiss genome of the influenza virus responsible for the pandemic of 1918-1920. For their study, the researchers used a more than 100-year-old virus taken from a formalin-fixed wet specimen sample in the Medical Collection of the Institute of Evolutionary Medicine at UZH. The virus came from an 18-year-old patient from Zurich who had died during the first wave of the pandemic in Switzerland and underwent autopsy in July 1918.
Three key adaptations in Swiss virus genome
“This is the first time we’ve had access to an influenza genome from the 1918-1920 pandemic in Switzerland. It opens up new insights into the dynamics of how the virus adapted in Europe at the start of the pandemic,” says last author Verena Schünemann. By comparing the Swiss genome with the few influenza virus genomes previously published from Germany and North America, the researchers were able to show that the Swiss strain already carried three key adaptations to humans that would persist in the virus population until the end of the pandemic.
Two of these mutations made the virus more resistant to an antiviral component in the human immune system – an important barrier against the transmissions of avian-like flu viruses from animals to humans. The third mutation concerned a protein in the virus’s membrane that improved its ability to bind to receptors in human cells, making the virus more resilient and more infectious.
New genome-sequencing method
Unlike adenoviruses, which cause common colds and are made up of stable DNA, influenza viruses carry their genetic information in the form of RNA, which degrades much faster. “Ancient RNA is only preserved over long periods under very specific conditions. That’s why we developed a new method to improve our ability to recover ancient RNA fragments from such specimens,” says Christian Urban, the study’s first author from UZH. This new method can now be used to reconstruct further genomes of ancient RNA viruses and enables researchers to verify the authenticity of the recovered RNA fragments.
Invaluable archives
For their study, the researchers worked hand in hand with UZH’s Medical Collection and the Berlin Museum of Medical History of the Charité University Hospital. “Medical collections are an invaluable archive for reconstructing ancient RNA virus genomes. However, the potential of these specimens remains underused,” says Frank Rühli, co-author of the study and head of the Institute of Evolutionary Medicine at UZH.
The researchers believe the results of their study will prove particularly important when it comes to tackling future pandemics. “A better understanding of the dynamics of how viruses adapt to humans during a pandemic over a long period of time enables us to develop models for future pandemics,” Verena Schünemann says. “Thanks to our interdisciplinary approach that combines historico-epidemiological and genetic transmission patterns, we can establish an evidence-based foundation for calculations,” adds Kaspar Staub, co-author from UZH. This will require further reconstructions of virus genomes as well as in-depth analyses that include longer intervals.
While the incidence of breast cancer is highest for white women, Black women are more likely to have early-onset or more aggressive subtypes of breast cancer, such as triple-negative breast cancer. Among women under 50, the disparity is even greater: young Black women have double the mortality rate of young white women.
Now, research from the University of Notre Dame is shedding light on biological factors that may play a role in this disparity. The study published in iScience found that a population of cells in breast tissues, dubbed PZP cells, send cues that prompt behavioural changes that could promote breast cancer growth.
Funded by the National Cancer Institute at the National Institutes of Health, the study set out to explore what biological differences in breast tissue could be related to early onset or aggressive breast cancers. Most breast cancers are carcinomas, or a type of cancer that develops from epithelial cells. In healthy tissue, epithelial cells form linings in the body and typically have strong adhesive properties and do not move.
The researchers focused on PZP cells as previous studies had shown that these cells are naturally and significantly higher in healthy breast tissues of women of African ancestry than in healthy breast tissues of women of European ancestry. While PZP cell levels are known to be elevated in breast cancer patients in general, their higher numbers in healthy, African ancestry tissues could hold clues to why early-onset or aggressive breast cancers are more likely to occur in Black women.
“The disparity in breast cancer mortality rates, particularly among women of African descent, is multifaceted. While socioeconomic factors and delayed diagnosis may be contributing factors, substantial emerging evidence suggests that biological and genetic differences between racial groups can also play a role,” said Crislyn D’Souza-Schorey, the Morris Pollard Professor of Biological Sciences at Notre Dame and corresponding author of the study.
The study showed how PZP cells produce factors that activate epithelial cells to become invasive, where they detach from their primary site and invade the surrounding tissue.
For example, a particular biological signaling protein known as AKT is often overactive in breast cancers. This study showed that PZP cells can activate the AKT protein in breast epithelial cells, which in part allows them to invade the surrounding environment. PZP cells also secrete and deposit certain proteins outside the cell that guide the movement of breast epithelial cells as they invade.
Overall, the results of the study emphasize multiple mechanisms by which PZP cells may influence the early stages of breast cancer progression and their potential contribution to disease burden.
The researchers also looked at how a targeted breast cancer drug, capivasertib, which inhibits the AKT protein, impacted PZP cells and found it markedly reduced the effects of the PZP cells on breast epithelial cells.
“It’s important to understand the biological and genetic differences within normal tissue as well as tumours among racial groups, as these variations could potentially influence treatment options and survival rates. And consequently, in planning biomarker studies, cancer screenings or clinical trials, inclusivity is important,” said D’Souza-Schorey, also an affiliate of Notre Dame’s Berthiaume Institute for Precision Health and Harper Cancer Research Institute.
Genetic ancestry is much more complicated than how people report their race and ethnicity. New research, using data from the National Institutes of Health’s (NIH) All of Us Research Program, finds that people who identify as being from the same race or ethnic group can have a wide range of genetic differences. The findings are reported in the American Journal of Human Genetics, a Cell Press journal.
As doctors and researchers learn more about how genetic variants influence the incidence and course of human diseases, the study of genetic ancestry has become increasingly important. This research is driving the field of precision medicine, which aims to develop individualised healthcare.
People whose ancestors came from the same part of the world are likely to have inherited the same genetic variants, but self-identified race and ethnicity don’t tell the whole story about a person’s ancestors. NIH’s All of Us Research Program was created in part to address this puzzle and to learn more about how genetic ancestry influences human health.
In the current study, the investigators looked at the DNA of more than 230 000 people who have volunteered to share their health information for All of Us. They compared it to other large DNA projects from around the world using a technique called principal component analysis (PCA) to visualize population structure and help identify genetic similarity between individuals and groups of people. This analysis showed that people in the US have very mixed ancestry, and their DNA doesn’t always match the race or ethnicity they write on forms. Instead of falling into clear groups based on race or ethnicity, people’s genetic backgrounds show gradients of variation across different US regions and states.
This is especially significant for people who identify as being of Hispanic or Latino origin. These people have a wide-ranging blend of ancestries from European, Native American, and African groups. Importantly, genetic ancestry among these people varies across the US in part because of historic migration patterns. For example, Hispanics/Latinos in the Northeast are more likely to have Caribbean (and thus African) ancestry, and those in the Southwest are more likely to have Mexican and Central American (and thus Native American) ancestry.
One specific discovery was that ancestry was significantly associated with body mass index (BMI) and height, even after adjusting for socio-economic differences. For example, West and Central African ancestries were associated with higher BMI, whereas East Africa ancestry was associated with lower BMI. There were similar findings showing that people with ancestral origins from different parts of Europe have different body measurements including height, with northern European ancestry associated with greater height and southern European ancestry associated with shorter height. This suggests that subcontinental differences in ancestry can have opposite effects on biological traits and diseases.
This finding suggests that the subcontinental differences in ancestry between individuals can have opposite effects on biological traits, diseases, and health outcomes, emphasizing the importance of not classifying individuals into broad ancestry groups such as African, European, or Asian. Doing this will help to make this research more accurate and will help to improve the field of precision medicine.
Researchers at Princeton University and the Simons Foundation have identified four clinically and biologically distinct subtypes of autism, marking a transformative step in understanding the condition’s genetic underpinnings and potential for personalised care.
Analysing data from over 5000 children in SPARK, an autism cohort study funded by the Simons Foundation, the researchers used a computational model to group individuals based on their combinations of traits. The team used a “person-centred” approach that considered a broad range of over 230 traits in each individual, from social interactions to repetitive behaviours to developmental milestones, rather than searching for genetic links to single traits.
This approach enabled the discovery of clinically relevant autism subtypes, which the researchers linked to distinct genetic profiles and developmental trajectories, offering new insights into the biology underlying autism. Their results were published July 9 in Nature Genetics.
The study defines four subtypes of autism: Social and Behavioural Challenges, Mixed ASD with Developmental Delay, Moderate Challenges, and Broadly Affected. Each subtype exhibits distinct developmental, medical, behavioural and psychiatric traits, and importantly, different patterns of genetic variation.
Individuals in the Social and Behavioural Challenges group show core autism traits, including social challenges and repetitive behaviours, but generally reach developmental milestones at a pace similar to children without autism. They also often experience co-occurring conditions like ADHD, anxiety, depression or obsessive-compulsive disorder alongside autism. One of the larger groups, this constitutes around 37% of the participants in the study.
The Mixed ASD with Developmental Delay group tends to reach developmental milestones, such as walking and talking, later than children without autism, but usually does not show signs of anxiety, depression or disruptive behaviours. “Mixed” refers to differences within this group with respect to repetitive behaviours and social challenges. This group represents approximately 19% of the participants.
Individuals with Moderate Challenges show core autism-related behaviours, but less strongly than those in the other groups, and usually reach developmental milestones on a similar track to those without autism. They generally do not experience co-occurring psychiatric conditions. Roughly 34% of participants fall into this category.
The Broadly Affected group faces more extreme and wide-ranging challenges, including developmental delays, social and communication difficulties, repetitive behaviours and co-occurring psychiatric conditions like anxiety, depression and mood dysregulation. This is the smallest group, accounting for around 10% of the participants.
“These findings are powerful because the classes represent different clinical presentations and outcomes, and critically we were able to connect them to distinct underlying biology,” said Aviya Litman, a PhD student at Princeton.
Distinct genetics behind the subtypes
For decades, autism researchers and clinicians have been seeking robust definitions of autism subtypes to aid in diagnosis and care. Autism is known to be highly heritable, with many implicated genes.
“While genetic testing is already part of the standard of care for people diagnosed with autism, thus far, this testing reveals variants that explain the autism of only about 20% of patients,” said co-author Jennifer Foss-Feig, a clinical psychologist at the Icahn School of Medicine at Mount Sinai and vice president and senior scientific officer at the Simons Foundation Autism Research Initiative (SFARI). This study takes an approach that differs from classic gene discovery efforts by identifying robust autism subtypes that are linked to distinct types of genetic mutations and affected biological pathways.
For example, children in the Broadly Affected group showed the highest proportion of damaging de novo mutations, while only the Mixed ASD with Developmental Delay group was more likely to carry rare inherited genetic variants. While children in both of these subtypes share some important traits like developmental delays and intellectual disability, these genetic differences suggest distinct mechanisms behind superficially similar clinical presentations.
“These findings point to specific hypotheses linking various pathways to different presentations of autism,” said Litman, referring to differences in biology between children with different autism subtypes.
Moreover, the researchers identified divergent biological processes affected in each subtype. “What we’re seeing is not just one biological story of autism, but multiple distinct narratives,” said Natalie Sauerwald, associate research scientist at the Flatiron Institute and co-lead author. “This helps explain why past genetic studies often fell short – it was like trying to solve a jigsaw puzzle without realising we were actually looking at multiple different puzzles mixed together. We couldn’t see the full picture, the genetic patterns, until we first separated individuals into subtypes.”
Autism biology unfolds on different timelines
The team also found that autism subtypes differ in the timing of genetic disruptions’ effects on brain development. Genes switch on and off at specific times, guiding different stages of development. While much of the genetic impact of autism was thought to occur before birth, in the Social and Behavioural Challenges subtype – which typically has substantial social and psychiatric challenges, no developmental delays, and a later diagnosis – mutations were found in genes that become active later in childhood. This suggests that, for these children, the biological mechanisms of autism may emerge after birth, aligning with their later clinical presentation.
“By integrating genetic and clinical data at scale, we can now begin to map the trajectory of autism from biological mechanisms to clinical presentation,” said co-author Chandra Theesfeld, senior academic research manager at the Lewis-Sigler Institute and Princeton Precision Health.
A paradigm shift for autism research
This study builds on more than a decade of autism genomics research led by Troyanskaya and collaborators. It is enabled by the close integration of interdisciplinary expertise in genomics, clinical psychology, molecular biology, computer science and modelling, and computational biology.
“The Princeton Precision Health initiative uses artificial intelligence and computational modelling to integrate across biological and clinical data,” said Jennifer Rexford, Princeton University provost and Gordon Y.S. Wu Professor in Engineering. “This initiative could not exist without the University’s charitable endowment. Our investments allow experts to collaborate across a range of disciplines to conduct transformative research that improves human health, including the potential for major advances in the diagnosis and treatment of autism made possible in this exciting project.”
“It’s a whole new paradigm, to provide these groups as a starting point for investigating the genetics of autism,” said Theesfeld. Instead of searching for a biological explanation that encompasses all individuals with autism, researchers can now investigate the distinct genetic and biological processes driving each subtype.
This shift could reshape both autism research and clinical care – helping clinicians anticipate different trajectories in diagnosis, development and treatment. “The ability to define biologically meaningful autism subtypes is foundational to realising the vision of precision medicine for neurodevelopmental conditions,” said Sauerwald.
While the current work defines four subtypes, “this doesn’t mean there are only four classes,” said Litman. “It means we now have a data-driven framework that shows there are at least four – and that they are meaningful in both the clinic and the genome.”
Looking ahead
Beyond its contributions to understanding autism subtypes and their underlying biology, the study offers a powerful framework for characterising other complex, heterogeneous conditions and finding clinically relevant disease subtypes. As Theesfeld put it: “This opens the door to countless new scientific and clinical discoveries.”
