The study, published in Science, found that the three-dimensional organization of the genome differs in certain brain cells from people with Alzheimer’s disease. Scientists from SCS’s Ray and Stephanie Lane Computational Biology Department, Pitt’s Department of Neurobiology, and collaborating institutions connected these changes in genome folding with shifts in gene activity and the organization of brain tissue.

To build this detailed picture, the team combined single-cell technology, spatial mapping of brain tissue, and a newly developed deep learning model.

Looking Beyond Amyloid and Tau

“Alzheimer’s disease cannot be understood one layer at a time,” said Jian Ma, the Ray and Stephanie Lane Professor of Computational Biology who led and supervised the study. “The genome’s 3D structure is a fundamental regulatory layer that helps to connect DNA sequence to gene activity. By integrating genome folding, cell state, and tissue context, we can move beyond cataloging disease-associated changes toward understanding how they fit together and which mechanisms to test next.”

DNA does not sit inside a cell as a simple straight strand. Instead, it folds into a complex three-dimensional structure that helps determine which genes are accessible and active. Changes in that physical organization can therefore influence how cells function.

The researchers examined postmortem samples from the prefrontal cortex, an area at the front of the brain. The tissue came from people with and without Alzheimer’s disease who had taken part in a long-term dementia study and later donated their brains for research.

The team used GAGE-seq, a technique that can measure both gene expression and three-dimensional genome contacts within the same individual cell. Those measurements were then combined with spatial transcriptomic maps, which preserve information about where gene activity occurs within intact brain tissue.

By bringing these datasets together, the researchers were able to connect the physical organization of the genome with gene regulation while also seeing where Alzheimer’s related molecular and cellular changes appeared within the surrounding tissue.

A New Layer of Alzheimer’s Biology

“Our study represents a major advance in understanding what goes wrong in Alzheimer’s disease,” said Hansruedi Mathys, assistant professor of neurobiology at Pitt’s Department of Neurobiology, who directed the Pitt arm of the study. “We know the classic hallmarks of Alzheimer’s disease – accumulation of amyloid-beta plaques and tau tangles – but our results establish higher-order chromatin alterations as a component of the molecular pathology associated with the disease, which currently affects seven million Americans, a number that continues to grow.”

Amyloid beta plaques and tau tangles are among the best known biological features of Alzheimer’s disease. The new findings suggest that changes in chromatin, the material made of DNA and associated proteins that packages the genome inside cells, should also be considered part of the disease’s molecular landscape.

AI Connects Genome Folding to Gene Activity

Another important part of the research was Hicformer, an artificial intelligence model developed to investigate how genome structure may influence cellular behavior. The model combines DNA sequence information with broad patterns of genome folding and detailed maps showing where different sections of DNA physically contact one another.

Using these inputs, Hicformer predicts gene activity across different types of cells. Xinyue Lu, a doctoral student in Computational Biology who co-led the research, described the system as a computational test bed that can be used to explore how changes in genome folding might alter gene activity.

“Measuring gene activity and genome folding in the same cell allows us to directly connect chromosome structure with disease-related gene programs,” said Yang Zhang, a project scientist in the Computational Biology Department who co-led the research. “Across several kinds of brain cells, this paired view revealed a consistent signature of 3D genome reorganization in Alzheimer’s disease and helped us prioritize regulatory regions for future mechanistic and therapeutic investigation.”

DNA Organization Becomes Less Distinct

The researchers identified several consistent differences in the genome architecture of cells from people with Alzheimer’s disease.

Large sections of the genome are normally organized into relatively distinct active and inactive regions known as compartments. In Alzheimer’s cells, those boundaries appeared less sharply defined. The researchers describe this pattern as “increased compartment mingling.”

Several kinds of brain cells also showed fewer interactions between nearby sections of the genome and more contacts between regions located farther apart. Cells with greater compartment mingling tended to have lower overall levels of gene activity.

The team also observed weaker interactions between genes and nearby regulatory elements that normally help control whether those genes are switched on or off. At the same time, some contacts across intermediate distances became stronger.

These structural differences were associated with reduced activity in programs involved in neurons and synapses, along with changes in metabolism and cellular stress responses. The researchers also found links to senescence-related programs in microglia, immune cells in the brain that play important roles in maintaining brain health and responding to damage.

Potential Clues for Future Alzheimer’s Treatments

When the researchers mapped these molecular changes across intact brain tissue, they found that the reorganization of the genome was connected not only to altered gene activity but also to differences in how brain cells were arranged within the tissue.

The results establish three-dimensional genome organization as another important layer of Alzheimer’s disease biology. They also provide researchers with a framework for testing which changes in genome architecture might directly contribute to the disease.

Future studies can now investigate whether particular structural changes help drive Alzheimer’s progression and whether any of the affected regulatory regions could eventually become targets for new therapies.

The research was supported by grants from the National Institutes of Health. Other CMU authors included doctoral students Shahul Alam and Shike Wang and postdoctoral research associate Junjie Tang. Other Pitt authors include doctoral students Alexander K. Kunisky and Jude Baroudi, post-baccalaureate research fellows Sahar and Sahel Ghorbanikalateh, and visiting scholar Shihan Wang. The team included researchers from the Broad Institute of MIT and Harvard; the University of California, Los Angeles; the University of Washington; and the Rush Alzheimer’s Disease Center.

Even the standard model of cosmology, known as ΛCDM (Lambda Cold Dark Matter), failed to win support from a majority of respondents. That result may reflect recent findings from the Dark Energy Spectroscopic Instrument (DESI), which suggested that dark energy could change over time. Such a possibility would conflict with the standard model, which assumes that dark energy remains constant.

And cosmology was far from the only area where physicists disagreed.

Standard Answers Fail To Win Broad Support

“The most striking result is how few of the ‘standard answers’ in fundamental physics command overwhelming support, with most falling short of a majority. The interesting point is not that physicists are confused. It is that the frontier is genuinely alive,” says Niayesh Afshordi, associate faculty member at Perimeter Institute and professor at the University of Waterloo.

Afshordi led the study with coauthor Phil Harper and the American Physical Society’s Physics Magazine.

Across the questions included in the survey, only two received majority agreement.

One concerned the Big Bang. Despite the way it is often portrayed in popular culture, 68% of the physicists surveyed said the Big Bang does not necessarily represent the beginning of time. Instead, the theory describes how the universe developed from an extremely hot and dense state. It does not, by itself, explain whether time had an absolute beginning.

The second point to cross the majority threshold was cosmic inflation. Just 51% agreed that the early universe experienced an extremely rapid period of expansion known as inflation.

Dark Matter Remains Wide Open

On many other major questions, the responses were much more divided.

Dark matter is one example. Only 17% favored the idea that dark matter is made of a yet undiscovered low-mass particle or particles. Another 12% supported modifications to the theory of gravity. The largest single group, at 21%, favored some combination of the many proposed explanations.

That spread of responses highlights how little consensus exists around one of the central mysteries of modern cosmology.

No Clear Winner for Quantum Gravity

Physicists were similarly divided over quantum gravity, the effort to develop a theory that can describe gravity within the framework of quantum mechanics.

String theory received the most support, but only 19% of respondents selected it as the most likely solution. Loop quantum gravity received 12%, while 18% favored the possibility that gravity cannot be quantized at all.

The result shows that even after decades of theoretical work, no single approach has emerged as the dominant answer.

Why Disagreement Could Be Good for Physics

So what does such widespread disagreement mean for the future of the field? Afshordi sees the lack of consensus as a sign of opportunity rather than failure.

“Scientific truth is not decided by a vote. But consensus, or its absence, tells us where the evidence feels settled and where researchers still see room for radically different ideas. In this sense, lack of consensus can be a clue. It marks places where better data, sharper theory, or new connections between subfields may be needed. In the eternal words of the Canadian singer and songwriter, Leonard Cohen: ‘There is a crack in everything, that’s how the light gets in.'”

Rather than suggesting that physicists have lost their way, the findings point to areas where major discoveries may still be possible. Some of the most fundamental questions about the universe remain open, leaving room for new observations, stronger theories, and unexpected ideas to reshape our understanding.

The survey results are described in an article published in Physics Magazine. An online dashboard also allows readers to explore the responses in greater detail.

The study, published recently in Nature Health, examined more than five years of posts from nearly 70,000 Reddit users. Two categories stood out as particularly deserving of further investigation: reproductive symptoms, including changes in menstrual cycles, and problems involving body temperature, such as chills and hot flashes.

The findings do not establish that the medications caused these symptoms. Instead, researchers say the massive collection of spontaneous patient reports may reveal signals worth examining more closely.

“Some of the side effects we found, like nausea, are well known, and that shows that the method is picking up a real signal,” says Sharath Chandra Guntuku, Research Associate Professor in Computer and Information Science (CIS) at Penn Engineering and the study’s senior author. “The underreported symptoms are leads that came from patients themselves, unprompted, and clinicians could potentially pay attention to them.”

What Patients Report Outside Clinical Trials

Clinical trials are designed to determine whether treatments work and to identify important safety problems, but they cannot necessarily capture every symptom that matters to patients once a medication is being used by a much larger population.

“Clinical trials generally identify the most dangerous side effects of drugs,” adds Lyle Ungar, Professor in CIS and a co-author on the study. “But they can fail to find what symptoms patients are most concerned about; even though social media is not necessarily representative, a large collection of posts may reflect additional concerns.”

The distinction is important. The study found associations in what people discussed online, not proof that GLP-1 drugs were responsible for those experiences.

“We can’t say that GLP-1s are actually causing these symptoms,” notes Neil Sehgal, the study’s first author and a doctoral student in CIS advised by Guntuku and Ungar. “But nearly 4% of the Reddit users in our sample reported menstrual irregularities, which would be even higher in a female-only sample. We think that’s a signal worth investigating.”

Using Social Media as an Early Health Signal

The idea of mining online conversations for clues about drug safety predates today’s AI boom. In 2011, Ungar participated in one of the earliest efforts to use material created by internet users to identify possible adverse effects from medications.

Social media can capture experiences that patients discuss with one another but may never formally report to a doctor, drug manufacturer, or regulator.

“Online patient communities work a lot like a neighborhood grapevine,” says Ungar. “People who are living with these medications are swapping notes with each other in real time, sharing experiences that rarely make it into a doctor’s office visit or an official report.”

Since then, online patient communities have expanded enormously. That has made social media a potentially valuable source for studying how medications affect people in everyday life, although gaining access to platform data has become more difficult.

Traditional clinical research remains essential, the researchers emphasize, but online conversations can provide information far more quickly when millions of people begin using a drug.

“Clinical trials are the gold standard, but by design, they are slow,” says Guntuku. “This is not a replacement for trials, but it can move much faster, and that speed matters when a drug goes from niche to mainstream almost overnight.”

AI Makes Massive Social Media Analysis Possible

One of the biggest obstacles has always been scale.

Guntuku describes the approach as “computational social listening,” which uses computational methods to identify patterns in large collections of online conversations about health.

Patients, however, rarely describe symptoms using standardized medical terminology. One person might describe feeling unusually cold, another might mention constant chills, while a clinician could categorize those experiences using a specific medical term.

Researchers therefore need a way to translate everyday language into standardized categories. One important reference is the Medical Dictionary for Regulatory Activities (MedDRA), which provides terminology widely used to classify medical conditions, symptoms, and adverse events.

Previously, connecting huge numbers of informal social media posts with standardized medical terminology required enormous amounts of work, limiting how much data researchers could realistically analyze.

Large language models such as GPT and Gemini are changing that equation by allowing researchers to process and categorize vast amounts of text more consistently and quickly.

“Large language models have made it possible to do this kind of analysis much faster with a level of standardization that could be difficult to achieve before,” says Sehgal.

Unexpected Symptoms Emerge From 400,000 Posts

The researchers stress that Reddit users do not represent the overall population of people taking GLP-1 drugs. Reddit users tend to be younger, are more likely to be male, and are disproportionately located in the United States.

Even with that limitation, the analysis produced a reassuring sign that the approach was detecting genuine patterns. Many of the symptoms discussed by Reddit users closely matched already known effects of semaglutide and tirzepatide.

About 44% of users included in the study described at least one side effect. Gastrointestinal problems were the most common, consistent with the nausea and other digestive issues already associated with these medications.

More intriguing were symptoms that appeared frequently enough to attract the researchers’ attention but may not be as well represented in current drug labels or conventional adverse event reports.

Nearly 4% of users who reported side effects described reproductive symptoms. These included changes in menstruation such as bleeding between periods, heavy bleeding, and irregular menstrual cycles.

Users also described changes involving body temperature, including chills, feeling unusually cold, hot flashes, and symptoms resembling a fever.

Fatigue was another notable finding. It was the second most frequently reported complaint in the Reddit data, even though relatively few clinical trials reported fatigue often enough for it to reach established reporting thresholds.

Why Menstrual and Temperature Changes Are Interesting

One possible reason these reports caught researchers’ attention involves the hypothalamus, a small but extremely important region of the brain. Among its many jobs, the hypothalamus helps regulate hunger, hormones, reproduction and body temperature.

“These drugs are thought to work by engaging part of the brain called the hypothalamus, which helps regulate a wide variety of hormones,” says Jena Shaw Tronieri, Senior Research Investigator at Penn’s Center for Weight and Eating Disorders and a co-author of the study. “That doesn’t mean the medications are necessarily causing these symptoms, but it could suggest that reports of menstrual changes and body temperature fluctuations are worth studying more systematically.”

The researchers are not proposing that this biological connection proves GLP-1 drugs are responsible. Instead, it offers another reason to test the patient-reported patterns more carefully through controlled research.

Turning Online Conversations Into Research Leads

For now, the team hopes the results encourage scientists and clinicians to pay closer attention to symptoms that patients repeatedly discuss online.

“They’re clearly on patients’ minds, and that’s worth paying attention to,” says Sehgal.

The researchers also want to broaden their analysis beyond Reddit and beyond English-language communities. Doing so could help determine whether the same patterns emerge among different groups of people and on different social media platforms.

“We don’t really know yet whether what we’re seeing on Reddit reflects the experience of GLP-1 users globally, or whether it’s particular to the kind of person who posts on Reddit in the United States,” Ungar says.

In the longer term, rapid AI analysis of online patient conversations could potentially become an early detection system for emerging health concerns involving drugs, supplements and wellness products.

That could be particularly valuable for substances that become popular online faster than conventional research can keep up. Loosely regulated or unregulated products, including injectable peptides, can spread quickly through communities on Reddit, TikTok and other platforms. Discussions among users may therefore provide some of the earliest indications of unexpected effects.

“The whole point of this kind of approach is that it can move quickly, and that’s exactly when it’s most valuable,” says Guntuku.

This study was conducted at the University of Pennsylvania School of Engineering and Applied Science. The authors report no outside funding. Tronieri reports receiving an investigator-initiated grant, on behalf of the University of Pennsylvania, from Novo Nordisk and receiving consulting fees from Currax Pharmaceuticals, LLC. The other authors report no conflicts of interest.

But new research from Northwestern Medicine suggests the scan may provide its greatest benefit to a narrower group of patients than previously assumed.

Researchers followed more than 6,000 adults for 10 years and found that, for the overall study population, adding a coronary artery calcium score provided only a small improvement over PREVENT, the American Heart Association’s primary cardiovascular risk calculator.

How Heart Calcium Scans Measure Risk

PREVENT estimates a person’s chance of developing cardiovascular disease over the next 10 or 30 years using commonly available health information (such as blood pressure, cholesterol, age, and sex).

A coronary artery calcium scan takes a different approach. The CT scan looks for calcium-containing plaque inside the coronary arteries. The resulting calcium score reflects the amount of detectable calcified plaque. In general, a higher score is associated with a greater risk of future cardiovascular disease.

The new findings suggest that calcium scores become particularly useful when PREVENT initially places someone in a borderline or intermediate risk category. Among these patients, the additional scan provided a clearer indication of who was more likely to develop cardiovascular disease and who was less likely.

“Coronary artery calcium scans are becoming more widely available and less expensive,” said study senior author Dr. Nilay Shah, assistant professor of medicine in the division of cardiology at Northwestern University Feinberg School of Medicine. “Our findings suggest that not everyone necessarily needs or would benefit from a coronary artery calcium scan for the purpose of predicting risk of heart attack and stroke.”

Shah said the findings also highlight potential downsides of using the scans when they are unlikely to change a patient’s treatment.

“Routinely using a calcium scan in people who are at low risk may result in unnecessary radiation exposure, testing and costs with unclear clinical benefits,” Shah added. “Using calcium scans in people who are at high risk is likely to result in unnecessary testing because these individuals are recommended to start a statin regardless of what the calcium scan shows.”

The study was published on Aug. 26 in JAMA.

Comparing Predictions With Real Outcomes

To investigate how much additional information calcium scans provide, Shah and his colleagues examined data from more than 6,000 adults ages 45 to 79 who participated in the Multi-Ethnic Study of Atherosclerosis.

At the beginning of the study, each participant received both a coronary artery calcium score and a PREVENT estimate of their likelihood of experiencing a cardiovascular event within the following 10 years.

Researchers then tracked what actually happened.

Over the next decade, 6% of the participants experienced either a heart attack or stroke. When the team compared predictions that included calcium scores with those based on PREVENT alone, the overall improvement was modest.

The model’s discrimination, which measures how effectively it distinguishes between people who will and will not experience a cardiovascular event, increased only slightly. It rose from 0.73 using PREVENT to 0.75 when calcium scores were included.

The picture changed, however, when researchers focused specifically on people whose initial PREVENT scores placed them in the borderline or intermediate risk categories (those initially deemed to have a 3% to 9% risk of heart disease within 10 years).

For this group, including the calcium score produced a more meaningful improvement in predicting future cardiovascular events.

“For patients at borderline risk, knowing their calcium score can help determine whether their risk is actually lower or higher than initially estimated, which can help guide treatment decisions,” Shah explained.

Finding the Patients Most Likely to Benefit

About 10% of U.S. adults age 30-79 have cardiovascular disease, which remains the nation’s leading cause of death. Yet many heart attacks, strokes and other cardiovascular events can be prevented, Shah said.

Accurately identifying a person’s risk can help doctors determine who is most likely to benefit from preventive treatments, including statins, which lower cholesterol and reduce cardiovascular risk.

“The findings help us understand how best to use the available tools to estimate someone’s risk of a heart attack or stroke. That provides more precise guidance for who is most likely to benefit from using a statin to help prevent heart disease,” Shah said.

Shah added that the results also support the clinical usefulness of the relatively new PREVENT calculator. Even without calcium scan results, PREVENT performed well at predicting cardiovascular risk.

Questions That Still Need Answers

The researchers said additional studies are needed to determine how much calcium scores improve PREVENT estimates in certain populations.

That includes higher risk groups such as South Asian and Filipino adults, Shah noted. More research is also needed in younger people because participants in this study were between ages 45 and 79 when the research began.

Other Northwestern co-authors are Xiaoning Huang, Lucia Petito, Norrina Allen, Dr. Philip Greenland and Dr. Sadiya Khan.

The study is titled, “Predictive Utility of Coronary Artery Calcium Added to the PREVENT Atherosclerotic Cardiovascular Disease Equations.” It was supported by the American Heart Association (grant 24CDA1266732) and the National Heart, Lung, and Blood Institute (contracts 75N92020D00001, HHSN268201500003I, N01-HC-95159, 75N92020D00005, N01-HC-95160, 75N92020D00002, N01-HC-95161, 75N92020D00003, N01-HC-95162, 75N92020D00006, N01-HC-95163, 75N92020D00004, N01-HC-95164, 75N92020D00007, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168, N01-HC-95169 and grant K23HL157766).

The advance, developed in the lab of Marko Lončar, Tiantsai Lin Professor of Electrical Engineering, could support the development of compact quantum networks built directly onto chips. It may also help enable hybrid quantum systems that combine several different kinds of quantum bits, or qubits.

The findings are published in Nature Physics. The experiments were led by Eliza Cornell, a recent Ph.D. graduate from the Lončar lab who is now a postdoctoral researcher at Boston University, and Zhujing Xu, a former postdoctoral scholar in Lončar’s group.

Using Sound to Carry Quantum Information

One promising approach to quantum networking uses the spin of an electron, associated with impurity in diamond, to store quantum information. Tiny packets of mechanical vibration called phonons can then serve as carriers that move information between qubit nodes.

The Lončar lab has played a major role in exploring the potential of these systems. Among its advances is a structure known as a phononic cavity, which traps mechanical vibrations so they can interact more strongly with the electron spin inside a qubit.

Phonons may have important advantages over light, which is more commonly used to move quantum information across chip-scale networks. At the same frequency, phonons have much shorter wavelengths than light, making it possible to build considerably smaller components and pack them more tightly together.

Phonons also interact readily with both solid-state spins and electromagnetic fields. That versatility makes them especially attractive for hybrid quantum technologies that bring together different types of qubits in a single system.

The Challenge of Preserving Quantum Memory

Using phonons, however, creates a major difficulty: protecting quantum memory.

Qubits are extremely sensitive to disturbances from their surroundings. To remain useful, they must preserve their quantum state long enough to store and process information. This ability is known as coherence.

Researchers often protect quantum memories from environmental interference using microwave pulses that separate, or decouple, the memory from surrounding noise. Those techniques do not work particularly well for qubits placed inside phononic cavities.

That limitation has made it difficult to achieve both strong interaction with phonons and long-lasting quantum memory in the same device.

“Dressed” Qubits Protected by Sound

The SEAS team addressed the problem by demonstrating what they describe as “all-mechanical coherence protection” for a silicon-vacancy spin in diamond.

Instead of relying on conventional microwave pulses, the researchers continuously applied a mechanical driving field made from phonons. This changed the qubit into a different kind of quantum state known as a “dressed” qubit.

The term “dressed” refers to the qubit effectively “wearing” a continuous acoustic field. In this condition, the qubit becomes less vulnerable to low-frequency noise from its surroundings.

Because the protection comes from a continuous mechanical field that is compatible with phononic cavities, the technique could operate inside the same structures that may eventually connect stationary nodes in quantum networks.

That gives phonons a potentially powerful dual role. They could transport quantum information between different parts of a network while simultaneously helping protect that information from environmental noise.

“We are solving two problems,” Cornell said. “We want the spin to have strong interaction with phonons, and we want the spin to have a long coherence time. Our paper demonstrates a method of extending the coherence time that is compatible with the silicon-vacancy center being in a cavity.”

Quantum Coherence Lasts About Three Times Longer

With the new method, the researchers increased the coherence time of the silicon-vacancy spin by roughly a factor of three.

The result demonstrates that continuous-wave mechanical noise suppression can extend quantum coherence in real devices, suggesting that microscopic sound waves could become an important tool for building more reliable and compact quantum systems.

“All-mechanical coherence protection and fast control of a spin qubit” was co-authored by Zhaoyou Wang, Hana K. Warner, Eliana Mann, Michael Haas, Smarak Maity, Graham Joe, Liang Jiang, Peter Rabl, and Benjamin Pingault.

The research received U.S. federal support from: the National Science Foundation under grant number EEC-1941583; the Air Force Office of Scientific Research under award numbers FA9550-23-1-0333 and FA9550-23-1-0338; and Q-NEXT, a U.S. Department of Energy Office of Science National Quantum Information Science Research Centers under award No. DE-FOA-0002253.

The work was performed in part at the Harvard Center for Nanoscale Systems, a member of the National Nanotechnology Infrastructure Network, which is supported by National Science Foundation award No. ECS-0335765.

The Harvard Office of Technology Development is actively pursuing patent protection and commercialization opportunities for the innovations arising from this research.