Amazing Science

New research from the University of Georgia reveals that artificial intelligence can be used to find planets outside of our solar system. The recent study demonstrated that machine learning can be used to find exoplanets, information that could reshape how scientists detect and identify new planets very far from Earth.

"One of the novel things about this is analyzing environments where planets are still forming," said Jason Terry, doctoral student in the UGA Franklin College of Arts and Sciences department of physics and astronomy and lead author on the study. "Machine learning has rarely been applied to the type of data we're using before, specifically for looking at systems that are still actively forming planets."

The first exoplanet was found in 1992, and though more than 5,000 are known to exist, those have been among the easiest for scientists to find. Exoplanets at the formation stage are difficult to see for two primary reasons. They are too far away, often hundreds of lights years from Earth, and the disks where they form are very thick, thicker than the distance of the Earth to the sun.

Data suggests the planets tend to be in the middle of these disks, conveying a signature of dust and gases kicked up by the planet. The research showed that artificial intelligence can help scientists overcome these difficulties. "This is a very exciting proof of concept," said Cassandra Hall, assistant professor of astrophysics, principal investigator of the Exoplanet and Planet Formation Research Group, and co-author on the study. "The power here is that we used exclusively synthetic telescope data generated by computer simulations to train this AI, and then applied it to real telescope data. This has never been done before in our field, and paves the way for a deluge of discoveries as James Webb Telescope data rolls in."

The James Webb Space Telescope, launched by NASA in 2021, has inaugurated a new level of infrared astronomy, bringing stunning new images and reams of data for scientists to analyze. It's just the latest iteration of the agency's quest to find exoplanets, scattered unevenly across the galaxy. The Nancy Grace Roman Observatory, a 2.4-meter survey telescope scheduled to launch in 2027 that will look for dark energy and exoplanets, will be the next major expansion in capability—and delivery of information and data—to comb through the universe for life.

he Webb telescope supplies the ability for scientists to look at exoplanetary systems in an extremely bright, high resolution, with the forming environments themselves a subject of great interest as they determine the resulting solar system. "The potential for good data is exploding, so it is a very exciting time for the field," Terry said.

A phenomenon called "skyglow" is stealing our celestial view, and the impact is far more dramatic when observed by the unaided human eye.

Light pollution has robbed eight out of 10 Americans, and nearly a third of all humans, of a view of our own home galaxy, according to new research out Thursday. The problem is something called "skyglow," which is the cumulative, diffuse brightening of the light sky from artificial light sources.  

A new study published in the journal Science uses crowdsourced data from a program called Globe at Night, which is run by the National Science Foundation-funded NOIRLab, a network of observatories. It finds that skyglow as perceived by human eyes is more of a problem compared with satellite measurements of artificial light on Earth. 

The study is the latest addition to a growing body of scientific literature on light pollution stretching back at least half a century.  By analyzing over 50,000 citizen scientist observations, the researchers found an increase in sky brightness of 9.6% over the past decade, compared to just two per cent per year measured by satellites. 

"At this rate of change, a child born in a location where 250 stars were visible would be able to see only around 100 by the time they turned 18," said the study's lead author Christopher Kyba, a researcher at the German Research Centre for Geosciences, in a statement.

The authors estimate that 80% of people in the US and 30% worldwide aren't able to see the ethereal arc of the Milky Way on a clear night. Part of the problem with what we're able to see with unaided eyes has to do with the types of lighting in use. "LED lights have a strong effect on our perception of sky brightness," said Kyba. "This could be one of the reasons behind the discrepancy between satellite measurements and the sky conditions reported by Globe at Night participants."

Satellites also have a hard time detecting light that is emitted horizontally from sources more prominent in cities like billboards or storefronts. 

Researchers reviewed data on the incidence of autoimmune diseases in the post-acute coronavirus disease 2019 period.

Background

COVID-19 has caused unprecedented morbidity and mortality across the globe. The immune protection conferred by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) vaccines and prior COVID-19 history, coupled with developing effective therapeutic agents, has reduced COVID-19-associated mortality. However, post-COVID-19 conditions continue to increase, particularly new-onset autoimmune diseases in COVID-19 convalescents. Autoimmune diseases have been reported following COVID-19 among adults; however, the prevalence and magnitude of the diseases, and the incidence risks among individuals infected with SARS-CoV-2, compared to uninfected individuals, are not well-characterized. Improving understanding of the impact of COVID-19 on the risk of developing post-acute COVID-19 complications, such as autoimmune disease, could aid in implementing preventive measures and initiating prompt treatment to prevent COVID-19-associated morbidities. The findings would also be highly relevant for future pandemics and to analyze the long-term protective effects of COVID-19 vaccines.

About the review

In the review, researchers presented findings of two large-scale cohort studies, evaluating the incidence of autoimmune diseases following acute SARS-CoV-2 infections using participants' electronic medical records.

Evidence on post-COVID autoimmune disease obtained before the studies

Studies comprising pediatric COVID-19 patients suffering from multisystem inflammatory syndrome among children (MIS-C) have indicated that COVID-19 results in dysregulated immunological responses. The clinical manifestations of MIS-C overlap with hyperinflammatory syndromes, including macrophage activation syndrome, Kawasaki disease, and toxic shock syndrome. The pathophysiology of COVID-19-associated immunological dysfunction includes molecular mimicry by SARS-CoV-2 proteins, multiorgan involvement due to the presence of angiotensin-converting enzyme 2 (ACE2) receptors, essential for SARS-CoV-2 entry into host cells, bystander immune cell activation, autoantigen release following SARS-CoV-2-induced tissue damage, superantigen-regulated lymphocyte activation, and epitope spread. Additionally, factors such as genetic susceptibility, age, and comorbidities may contribute to COVID-19 pathogenesis. A previous study comparing the immune responses in SARS-CoV-2 infection and autoimmune disease reported that tissue damage in both conditions is mainly immune-regulated, demonstrated by the presence of anti-nuclear antibodies, anti-Ro/SSA antibodies, and lupus anticoagulant cold agglutinins in both conditions. A retrospective analysis was performed from January 31, 2020, to June 30, 2021, using the Clinical Practice Research Datalink Aurum database, comprising data from 458,147 and 1,818,929 SARS-CoV-2-infected and uninfected adults, respectively, residing in England. A study preprint reported that the incidence of inflammatory bowel disease, psoriasis, and type 1 diabetes was significantly related to COVID-19.

Findings of the two cohort studies assessed post-COVID autoimmune disease

Chang et al. utilized the TriNetX health research network, comprising six million adults across 48 health organizations globally. The propensity score-matched SARS-CoV-2-infected and uninfected groups comprised 887,455 COVID-19 vaccinees each. Post-COVID-19 autoimmune disease incidence was evaluated between January 1, 2020, and December 31, 2021. At six months of follow-up, autoimmune disease incidence was significantly greater in SARS-CoV-2-infected individuals, including systemic lupus erythematosus, rheumatoid arthritis, vasculitis, type 1 diabetes, and inflammatory bowel disease, with adjusted hazards ratios (aHR) of 2.99,2.98, 1.96, 2.68, and 1.78, respectively. Post-COVID autoimmune disease risk was consistent across ages. Tesch et al. conducted a study, yet to be peer-reviewed, to evaluate autoimmune disease risks among 640,701 unvaccinated individuals having polymerase chain reaction (PCR)-confirmed SARS-CoV-2 infections in 2020. The findings showed that COVID-19 patients were 43.0% more likely to acquire autoimmune conditions, within three to 15.0 months of infection, compared to over one million age- and sex-matched uninfected individuals. Among autoimmune diseases, the incidence rate ratio (IRR) values were the highest for vasculitis. Further, among those with prior history of autoimmune diseases, COVID-19 resulted in a 23.0% higher risk of another autoimmune disease. Generally, most autoimmune diseases were not COVID-19-specific; however, an essential aspect of SARS-CoV-2 infection was a considerable rise in the incidence and spectrum of autoimmune diseases following acute COVID-19.

Conclusions

Based on the review findings, COVID-19 elevates autoimmune disease risks. Due to the retrospective study designs, the two cohort studies could not provide a causal association between COVID-19 and autoimmune disease development. However, the temporal relationship with COVID-19 history provides reliable and compelling evidence that COVID-19 is associated with elevated autoimmune disease development risks. Health governments and authorities must conduct future studies on the topic to obtain national data and increase the generalizability of the findings. The definitive mechanisms, including genetic and epigenetic predilection, underlying the association and related pathophysiology are not completely understood. However, further research must be performed using particular gene-deficient experimental animals, bioinformatic analysis, and biological approaches, e.g., analyzing transcriptomic pandemic data to obtain genomic signatures, and host responses to viral triggers must be evaluated.

Research cited published (April 12, 2023) In Nature Reviews Rheumatology:

https://www.nature.com/articles/s41584-023-00964-y 

Aging, or senescent cells, which stop dividing but don’t die, can accumulate in the body over the years and fuel chronic inflammation that contributes to conditions such as cancer and degenerative disorders. In mice, eliminating senescent cells from aging tissues can restore tissue balance and lead to an increased healthy lifespan. Now a team led by investigators at Massachusetts General Hospital (MGH) has found that the immune response to a virus that is ubiquitously present in human tissues can detect and eliminate senescent cells in the skin.

For the study, which is published in Cell, the scientists analyzed young and old human skin samples to learn more about the clearance of senescent cells in human tissue. The researchers found more senescent cells in the old skin compared with young skin samples. However, in the samples from old individuals, the number of senescent cells did not increase as individuals got progressively older, suggesting that some type of mechanism kicks in to keep them in check. Experiments suggested that once a person becomes elderly, certain immune cells called killer CD4+ T cells are responsible for keeping senescent cells from increasing. Indeed, higher numbers of killer CD4+ T cells in tissue samples were associated with reduced numbers of senescent cells in old skin. When they assessed how killer CD4+ T cells keep senescent cells in check, the researchers found that aging skin cells express a protein, or antigen, produced by human cytomegalovirus, a pervasive herpesvirus that establishes lifelong latent infection in most humans without any symptoms. By expressing this protein, senescent cells become targets for attack by killer CD4+ T cells.

“Our study has revealed that immune responses to human cytomegalovirus contribute to maintaining the balance of aging organs,” says senior author Shawn Demehri, MD, PhD, director of the High Risk Skin Cancer Clinic at MGH and an associate professor of Dermatology at Harvard Medical School. “Most of us are infected with human cytomegalovirus, and our immune system has evolved to eliminate cells, including senescent cells, that upregulate the expression of cytomegalovirus antigens.” These findings, which highlight a beneficial function of viruses living in our body, could have a variety of clinical applications. “Our research enables a new therapeutic approach to eliminate aging cells by boosting the anti-viral immune response,” says Demehri. “We are interested in utilizing the immune response to cytomegalovirus as a therapy to eliminate senescent cells in diseases like cancer, fibrosis and degenerative diseases.” Demehri notes that the work may also lead to advances in cosmetic dermatology, for example in the development of new treatments to make skin look younger.

Most people are still not eligible for a second booster, but the FDA will revisit the topic after an expert panel meets in June. The Food and Drug Administration on Tuesday approved a second omicron booster from Moderna and Pfizer for people over the age of 65 and immunocompromised individuals.

Why it matters: An additional dose of the bivalent vaccine can help high-risk individuals with waning immunity to COVID, the FDA said.

• Most people who have received a single shot of the bivalent vaccine aren't eligible for a second dose, but the FDA could revise criteria after an expert panel meets to discuss the topic in June.

Details: Tuesday's decision means those over 65 can get a second booster no sooner than four months after the first.

• Immunocompromised people may receive the second shot at least two months after receiving the first dose "and additional doses may be administered at the discretion of, and at intervals determined by, their health care provider," the FDA said.
• The agency made additional changes, saying that unvaccinated people may receive a single dose of the bivalent vaccine as their first shot, rather than multiple doses of the original monovalent vaccine.
• Children between six months and five years old who have received up to three doses of the monovalent vaccine may receive a bivalent vaccine shot, but the number of doses they receive "will depend on the vaccine and their vaccination history," per the FDA.
• The monovalent shots from Moderna and Pfizer are no longer authorized for use in the U.S.

What they're saying: "At this stage of the pandemic, data support simplifying the use of the authorized mRNA bivalent COVID-19 vaccines and the agency believes that this approach will help encourage future vaccination," said Peter Marks, director of FDA's Center for Biologics and Research.

• "Evidence is now available that most of the U.S. population 5 years of age and older has antibodies to SARS-CoV-2 ... either from vaccination or infection that can serve as a foundation for the protection provided by the bivalent vaccines," Marks added.

Catch up fast: An FDA advisory committee in January unanimously recommended that the U.S. replace initial doses of the original COVID monovalent shots with bivalent ones to target omicron subvariants, directing Pfizer, Moderna and Novavax to focus on bivalent vaccines.

Astronomers have picked up eight mysterious radio signals that could be coming from aliens sending messages with technology more advanced than our own. The electromagnetic waves were detected using state of the art AI, or deep learning. The signals were sourced to areas surrounding five ‘nearby’ stars 30 to 90 light years away. The pulses were ‘hiding in plain sight’ among a huge number of recordings from more than six years ago.

An international team developed a computer algorithm to analyze the unimaginably large amount of information in more detail. Lead author Peter Ma, an undergraduate at the University of Toronto, said his team searched through 150 TB of data from 820 nearby stars. One terabyte could hold 1,000 copies of the Encyclopedia Brittanica.

“The dataset had previously been searched through in 2017 by classical techniques but labeled as devoid of interesting signals.”

It was collected by the Green Bank Telescope in West Virginia, which is bigger than the Statue of Liberty and part of the Breakthrough Listen project aimed at identifying extra terrestrial activity. No ‘targets of interest’ were originally indicated. But the new neural network found this to be far from the case.

Manual re-examination also confirmed the findings shared several key characteristics. The signals were narrow band, meaning they had a small spectral width of just a few Hz (Hertz). Natural phenomena tend to be broadband.

“The key issue with any techno-signature search is looking through this huge haystack of signals to find the needle that might be a transmission from an alien world,” explained Dr. Steve Croft, a California astrophysicist with the Breakthrough Listen team (and one of Ma’s research advisors). “Peter’s algorithm gives us a more effective way to filter the haystack and find signals that have the characteristics we expect from techno-signatures.” Furthermore, the readings, reported in the journal Nature Astronomy, were ‘sloped’, indicating acceleration. They also appeared only when the instrument focused on a specific celestial source, disappearing when it pointed away. Radio is a great way to send interstellar information. It passes through dust and gas at the speed of light—20,000 times faster than our best rockets.

Many SETI (search for extraterrestrial intelligence) efforts use antennas to eavesdrop on any signals aliens might be transmitting. “These results dramatically illustrate the power of applying modern machine learning and computer vision methods to data challenges in astronomy, resulting in both new detections and higher performance,” said co-author Dr. Cherry Ng, of the French National Centre for Scientific Research in Paris. “Application of these techniques at scale will be transformational for radio techno-signature science.”

The researchers are now planning to deploy the algorithm on the SETI Institute’s COSMIC tool in New Mexico, where Jodie Foster heard an alien signal in the 1997 movie Contact. It’s been surveying 40 million stars for ‘techno-signatures’ Since SETI experiments began in 1960 with Frank Drake’s Project Ozma at the same Green Bank Observatory used in this latest work, technological advances have enabled researchers to collect more data than ever. The massive volume requires supercomputers and AI to break new ground in the quest to answer the old fundamental question, ‘Are we alone?’.

Researchers have developed a computer vision-based approach to sharpening images captured by ground-based telescopes.

While vital to humanity’s existence on Earth, the planet’s atmosphere is a major nuisance for astronomers trying to learn more about what’s beyond Earth. The cosmos would be much easier to study without the pesky atmosphere. Researchers at Northwestern University in Chicago and Tsinghua University in Beijing have unveiled a new artificial intelligence-powered approach to cleaning up images captured by ground-based telescopes.

The primary culprit behind blurry images is shifting pockets of air in the atmosphere. These air pockets can obfuscate the shapes of celestial bodies and can also affect the measurements scientists can make of distant objects, leading to incorrect conclusions.

Space-based telescopes, such as Hubble and the James Webb Space Telescope, avoid this issue altogether by leaving Earth’s atmosphere. However, space telescopes are monumentally expensive and always in high demand by astronomers.

Ground-based telescopes are easier to build, more affordable, and much simpler to fix when issues arise. To deal with the blur caused by the atmosphere, ground-based telescopes are often built in arid climates and at high altitudes. Even still, blur must be dealt with.

The light emanating from distant stars, planets, and galaxies travels a very long way to reach Earth. The light’s journey may encounter different gravitational effects, which can warp the light as observed from Earth. However, the light must also travel through Earth’s atmosphere before it reaches ground-based telescopes.

“Even clear night skies still contain moving air that affects light passing through it. That’s why stars twinkle and why the best ground-based telescopes are located at high altitudes where the atmosphere is thinnest,” Northwestern explains. Alexander describes the effect as “a bit like looking up from the bottom of a swimming pool.” The atmosphere is much less dense than water, but the concept is similar. The resulting blur is a significant problem for astrophysicists analyzing images for important cosmological data. The apparent shape of galaxies sheds light on large-scale cosmological structures and their gravitational effects.

“Slight differences in shape can tell us about gravity in the universe. These differences are already difficult to detect. If you look at an image from a ground-based telescope, a shape might be warped. It’s hard to know if that’s because of a gravitational effect or the atmosphere,” Alexander says. If an image is blurry, it’s hard to conclude much about the shape of what’s being observed. By removing the blur precisely and mathematically, the true nature of distant objects can be measured and analyzed.

Scientists use numerous software-based approaches to process images to make them sharper and more useful. However, the new computer-vision algorithm, first seen on Space, approach promises better and faster results. The algorithm the team built has been explicitly adapted for processing astronomical images captured by ground-based telescopes, the first time such AI technology has been used for this purpose. The team has trained the AI on data simulated to match the Vera C. Rubin Observatory, which will house the world’s largest digital camera when it opens next year in Chile. The new algorithm will be immediately compatible as soon as the telescope is ready for operations. It’s like the impressive observatory has received an upgrade before it even starts its stargazing.

“Photography’s goal is often to get a pretty, nice-looking image,” says Northwestern’s Emma Alexander, the study’s senior author. “But astronomical images are used for science. By cleaning up images in the right way, we can get more accurate data. The algorithm removes the atmosphere computationally, enabling physicists to obtain better scientific measurements. At the end of the day, the images do look better as well.”

Alexander is an assistant professor of computer science at Northwestern’s McCormick School of Engineering, where she leads the Bio Inspired Vision Lab. The new study was co-authored by Tianao Li, an undergraduate student in electrical engineering at Tsinghua University and a research intern in Alexander’s lab. The team’s optimization algorithm, built upon a deep learning network trained on astronomical images, produces images with 38.6 percent less error than classic methods for removing blur and 7.4 percent less error compared to modern methods. That may not seem like a huge difference, but it is in the realm of astronomy. Relatively small improvements in accuracy and efficiency can have massive, far-reaching benefits to cutting-edge astronomy.

An international team of scientists, led by a researcher at The University of Manchester, have developed a novel AI (artificial intelligence) approach to extract technical astronomy terminology into simple understandable English in their recent publication.

The new research is a result of the international RGZ EMU (Radio Galaxy Zoo EMU) collaboration and is transitioning radio astronomy language from specific terms, such as FRI (Fanaroff-Riley Type 1), to plain English terms such as “hourglass” or “traces host galaxy”.

In astronomy, technical terminology is used to describe specific ideas in efficient ways that are easily understandable amongst professional astronomers. However, this same terminology can also become a barrier to including non-experts in the conversation.

The RGZ EMU collaboration is building a project on the Zooniverse citizen science platform, which asks the public for help in describing and categorizing galaxies imaged through a radio telescope.

Modern astronomy projects collect so much data that it is often impossible for scientists to look at it all by themselves, and a computer analysis can still miss interesting things easily spotted by the human eye. 

Micah Bowles, Lead author and RGZ EMU data scientist, said: “Using AI to make scientific language more accessible is helping us share science with everyone. With the plain English terms we derived, the public can engage with modern astronomy research like never before and experience all the amazing science being done around the world.” 

With chatbots like ChatGPT making a splash, machine learning is playing an increasingly prominent role in our lives. For many of us, it’s been a mixed bag. We rejoice when our Spotify For You playlist finds us a new jam, but groan as we scroll through a slew of targeted ads on our Instagram feeds.

Machine learning is also changing many fields that may seem surprising. One example is my discipline, ornithology, the study of birds. It isn’t just solving some of the biggest challenges associated with studying bird migration; more broadly, machine learning is expanding the ways in which people engage with birds. As spring migration picks up, here’s a look at how machine learning is influencing ways to research birds and, ultimately, to protect them.

The challenge of conserving migratory birds

Most birds in the Western Hemisphere migrate twice a year, flying over entire continents between their breeding and nonbreeding grounds. While these journeys are awe-inspiring, they expose birds to many hazards en route. These include extreme weather, food shortages and light pollution that can attract birds and cause them to collide with buildings. Our ability to protect migratory birds is only as good as the science that tells us where they go. And that science has come a long way.

Banding to track bird migration

In 1920, the U.S. Geological Survey launched the Bird Banding Laboratory, spearheading an effort to put bands with unique markers on birds, then recapture the birds in new places to figure out where they traveled. Today researchers can deploy a variety of lightweight tracking tags on birds to discover their migration routes. These tools have uncovered the spatial patterns of where and when birds of many species migrate. However, tracking birds has limitations. For one thing, over 4 billion birds migrate across the continent every year. Even with increasingly affordable equipment, the number of birds that we track is a drop in the bucket. And even within a species, migratory behavior may vary across sexes or populations. Further, tracking data tells us where birds have been, but it doesn’t necessarily tell us where they’re going. Migration is dynamic, and the climates and landscapes that birds fly through are constantly changing. That means it’s crucial to be able to predict their movements.

Using machine learning to forecast bird migration

This is where machine learning comes in. Machine learning is a subfield of artificial intelligence that gives computers the ability to learn tasks or associations without explicitly being programmed. We use it to train algorithms that tackle various tasks, from forecasting weather to predicting March Madness upsets.

But applying machine learning requires data. And the more data the better. Luckily, scientists have inadvertently compiled decades of data on migrating birds through the Next Generation Weather Radar system. This network, known as NEXRAD, measures weather dynamics and helps predict future weather events. But it also picks up signals from birds as they fly through the atmosphere.

A research team led by gene editing pioneer David R. Liu, PhD, reports the application of base editing technology to develop a one-time treatment for spinal muscular atrophy (SMA), showing promising effectiveness in cell and mouse models. Liu, a professor at Harvard University, the Broad Institute, and an investigator with the Howard Hughes Medical Institute, together with 16 co-authors, published their findings today in Science, in a report entitled, “Base editing rescue of spinal muscular atrophy in cells and in mice.”

The paper is the first one published in a peer-reviewed journal that details a base editing approach to treating SMA—the leading genetic cause of infant mortality. Two months ago, Benjamin P. Kleinstiver, PhD, assistant professor in the Center for Genomic Medicine at Massachusetts General Hospital (MGH) and Harvard Medical School, and 12 co-authors posted a related preprint on bioRxiv detailing successful results from their study to optimize and implement a base editing treatment for SMA. In SMA, patients have either a loss of, or mutations in, the SMN1 gene. The result is a lack of its corresponding protein SMN, which is critical to cellular functions that include those required for embryo development. Liu and co-authors detailed how they restored SMN protein abundance to normal levels and rescued disease phenotypes in cell and mouse models of SMA by developing one-time genome editing approaches targeting another gene, called SMN2.

SMN2 differs from SMN1 at just one position—a single C-to-T mutation in exon 7. The result is a defective truncated SMN protein that fails to fully compensate for the loss of SMN1, but which provides just enough SMN protein activity to allow babies to be born with SMA. In the most common SMA cases (type 1), those babies typically live less than two years without treatment. Liu and colleagues developed a customized base editor (out of more than 40 tested) that directly converts SMN2 to the same sequence as SMN1 by reversing the C-to-T mutation that occurred during the duplication of SMN1 to SMN2.

The investigators reported that their strategy converted SMN2 to SMN1 with near-ideal editing efficiency of 99% in cells, fully restoring SMN protein levels ~40-fold to normal levels, while generating minimal off-target edits. “Because all SMA patients have SMN2—otherwise loss of SMN1 would not allow them to be born—this base editing approach to treating SMA also should be applicable to all SMA patients, regardless of the specific mutation that caused their SMN1 loss,” Liu said.