A team led by bioengineers at the University of California San Diego has developed a genome-scale reference map that details how individual genes control the functions and identities of human stem cells. This open-access resource could help researchers build virtual cell models for complex diseases, as well as design patient-specific treatments for these diseases.

The study was published in Nature Biotechnology.

“The result is a kind of reference atlas; it’s a way to look up what perturbing almost any gene does to a stem cell’s behavior, measured here as the impact on its whole transcriptome,” said study senior author Prashant Mali, professor in the Shu Chien-Gene Lay Department of Bioengineering at the UC San Diego Jacobs School of Engineering.

This is the first genome-scale map of gene function in human induced pluripotent stem cells, which are adult cells that have been reprogrammed back into an embryonic-like state and can turn into any type of cell in the body, such as muscle, heart, skin or bone. Such a reference map is needed because the vast majority of what human genes actually do inside these cells remains a mystery.

To build this comprehensive reference map, the team used CRISPR technology to systematically switch off 11,692 expressed genes one by one and then measured the effect on cellular transcriptomes across more than 2.5 million single cells.

By compiling these data, the researchers grouped related genes and cellular components based on shared molecular traits and functions. This allowed them to isolate previously hidden metabolic and self-renewal genes. The map enabled the researchers to uncover previously unrecognized cell regulators and confirm their roles experimentally. For example, they identified a specific gene, called DBR1, as the main regulator of RNA editing—specifically, the conversion of adenosine to inosine.

The team envisions that this open-access map could serve as a resource for streamlining and accelerating biomedical research.

“The map we generated works as a hypothesis engine—it’s a starting point for what a given gene does and which genes might be worth pursuing as targets to drive differentiation into cell states of interest,” said study co-first author Yesh Doctor, a bioengineering Ph.D. student in Mali’s lab. “Scientists can use it to look up the functions of genes and build hypotheses about them instead of having to run the experiments themselves.”

“We are grateful for the support of the NIH, especially the Bridge2AI program and NHGRI,” Mali said. “These comprehensive, genome-scale screens enable generation of reference maps that are not only invaluable for basic science discovery but also important resources for powering future computational and AI tools for genotype-phenotype prediction, one of the central pursuits in genetics research.”

The map is accessible here.

Long before scientists began studying them in the lab, mind-altering substances were already being gathered from plants, fungi and even animals for use in rituals, healing practices and mental health treatment. Researchers at the Weizmann Institute of Science have now managed to bring together in a single organism five psychedelic substances that in nature are scattered across the tree of life.

After uncovering how plants naturally produce one of the best-known psychedelic compounds, DMT, they were able to reengineer that process step by step inside a model plant—along with four other psychedelics. The result is what amounts to a biological factory that could, in the future, be used to simultaneously produce multiple psychedelic molecules, including some that do not naturally occur in plants.

The study was led by Dr. Paula (Shirley) Berman, who worked at the time in Prof. Asaph Aharoni’s lab in Weizmann’s Plant and Environmental Sciences Department; she is now a principal investigator at the Agricultural Research Organization–Volcani Institute. The findings were published in Science Advances.

5 psychedelics, 3 kingdoms

The five compounds in the study—all well-known psychedelics—come from three different kingdoms of life. The plant kingdom contributed DMT, the brain-active component of ayahuasca, a ceremonial hallucinogenic brew long used in shamanic Amazonian rituals for spiritual healing.

The researchers derived DMT from several plant sources, including the leaves of a woody shrub from the coffee family, native to the Amazon rainforest, and the bark of an acacia species native to the Australian outback.

From the kingdom of fungi, they took psilocybin and psilocin—the compounds responsible for the effects of “magic mushrooms,” with psilocybin once having been central to Aztec ceremonies. Representing the animal kingdom was the Sonoran Desert toad; it has glands on its head and skin that release a milky defensive secretion when it is stressed.

This secretion contains bufotenin, as well as a more potent relative of DMT called 5-MeO-DMT, known to induce distinct psychedelic experiences—a fact well-known to those who have sought out the toad with the express purpose of licking it.

Despite their diverse origins, all five compounds belong to the same chemical family and share the same starting point: tryptophan, a common amino acid found in all living organisms. This is also the starting point the human body uses to produce serotonin, a neurotransmitter involved in regulating mood and well-being. That shared origin helps explain why psychedelics act on the same receptors in the brain as serotonin.

Rebuilding DMT inside tobacco

“At the heart of the study was the challenge of making DMT,” Aharoni explains.

Although scientists had previously mapped the general route of DMT production in nature, the exact genes and enzymes responsible were still unknown, and identifying the complete biosynthetic DMT pathway remained elusive. The researchers began by identifying the key genes, particularly those encoding the enzymes that drive each step of the pathway.

They then inserted these genes into a model plant—Nicotiana benthamiana, a tobacco relative widely used in research—effectively teaching it to produce DMT. Within days, the engineered plant began generating the compound.

When the scientists produced the other four psychedelics individually in separate tobacco plants, one of them—5-MeO-DMT—was manufactured in surprisingly low amounts.

To address this, the team collaborated with Prof. Sarel Fleishman and Dr. Olga Khersonsky of Weizmann’s Biomolecular Sciences Department, experts in protein design. They identified a subtle problem: a molecule that did not fit well into the active site of one of the enzymes. By changing a single building block—one amino acid—in the enzyme’s structure, they improved the fit.

The result was dramatic. “We mutated one amino acid in the sequence and got a 40-fold increase in the production of 5-MeO-DMT,” Berman says.

One plant, competing pathways

The scientists then introduced genes for the five compounds into the same plant. The system worked. A single plant was able to produce all five psychedelics: plant-origin DMT; fungus-origin psilocin and psilocybin; and animal-origin bufotenin and 5-MeO-DMT.

“In effect, we created a kind of biological ‘cocktail’—not by mixing substances externally, but by combining the underlying pathways inside one organism,” Aharoni says.

At the same time, the experiment revealed an important limitation. When multiple pathways were activated at once, they began to compete for the same starting material. In biological terms, the system reached a bottleneck, and production efficiency dropped.

Beyond nature’s own chemistry

Finally, the team pushed the system beyond what occurs in nature. By adding bacterial enzymes, they produced modified psychedelic molecules carrying chlorine or bromine atoms in specific positions—something that evolution had apparently left out of the plant’s job description but might prove therapeutically valuable.

Several such molecules have already shown intriguing biological activity, including antidepressant-like effects, as part of the growing search for new treatments for disorders such as depression, anxiety, PTSD and addiction.

A faster, less extractive source

The research points toward new ways of producing psychedelic compounds. Many are currently obtained from slow-growing plants, rare fungi or animal sources, often raising ecological and ethical concerns. The Sonoran Desert toad, for example, is increasingly threatened by habitat loss and overcollection. Plants used for ayahuasca are also under growing pressure because of land loss and rising demand.

Producing these molecules in fast-growing laboratory plants could provide a more sustainable alternative, reducing the need to harvest vulnerable species while making production more efficient and scalable. Plants are grown, the genes are introduced, and within about a week, measurable amounts of the psychedelic can be extracted.

More available molecules mean more opportunities for research. One open question is why plants produce these compounds in the first place. Psychedelic molecules did not evolve so humans could “trip,” or to treat anxiety or depression; they likely serve ecological roles, such as defense or interactions with microbes and insects. By engineering plants to produce them in controlled settings, researchers can begin to study these possibilities directly.

“If we can move these pathways into a model plant that grows quickly and is easy to manipulate, we can start asking what these compounds actually do for the plant,” Berman explains. Researchers can examine how they affect the plant’s defenses or whether they influence its growth or stress responses.

From ayahuasca to edible doses

The scientists are also exploring the possibility of engineering a plant that produces the full ayahuasca mixture. In traditional preparations, DMT is combined with another compound that allows the brew to be active when swallowed.

In the Amazon, this is achieved by mixing leaves containing DMT with twigs bearing another substance that facilitates DMT’s absorption from the digestive tract. Scientists now aim to create a single plant that would contain both components.

Another potential direction involves producing therapeutic psychedelics in edible plants, so the substances could be consumed in carefully regulated doses.

All in all, the Weizmann study is not only about psychedelic compounds. It points to a broader shift in the relationship between plant biology and drug development—one in which plants are no longer just sources of rare molecules, but living platforms for studying, reshaping and potentially producing the next generation of psychiatric treatments.

Virtual fences could make managing grazing livestock on farms more flexible and more efficient while improving animal welfare. A new study by the University of Göttingen shows that virtual fences trigger behavior in cattle similar to that caused by conventional electric fences, in terms of how they move around the field. This finding puts commonly expressed concerns about animal welfare into perspective. The results were published in the journal Animal.

In previous research on this theme, cattle wore collars that emitted acoustic signals—and, where necessary, electrical pulses—when they approached the boundary. During the learning process, the animals associated the warning tone with the unpleasant stimulus and subsequently mostly respected the virtually defined boundary simply in response to the noise. As previous analyses of the experiment had not revealed any significant behavioral differences between animals in virtual and conventionally fenced enclosures, the researchers decided to take a closer look in their current study.

To map the distance to the fence—and identify any effect on the cattle’s behavior that might occur only in the vicinity of the virtual boundary—they used GPS to assign the movement data of the 31 cows to two different zones (peripheral zone and center of the pasture) and compared them.

The key finding was that it was not the type of fence that made the difference, but the boundary itself—regardless of whether it was visible or not. Regardless of the system used, the animals were less likely to be found at the edges of the field, moved more slowly there, and tended to use the center of the pasture. Virtual fences also led to a more even distribution of the animals across the area.

“Our findings show that it is not the type of fence that is the deciding factor, but rather the animals’ perception of the boundary of the pasture,” says lead author Dr. Natascha Grinnell at Göttingen University’s Institute of Grassland Science. “Virtual fences are respected by cattle just as reliably as conventional electric fences and are not fundamentally more problematic from an animal welfare perspective. This opens up new opportunities for farmers to manage grazing in a modern and flexible way.”

Researchers from the University of Göttingen will also present their findings at the ‘Virtual Fencing’ field day on Monday, July 6, in Alt Madlitz, Brandenburg.

Rutgers Office for Research (OfR) leaders collaborated with researchers around the world to develop an artificial intelligence (AI) program that has the potential to revolutionize lab research.

The international collaboration of researchers that built the program presented its findings in a paper published in Lab Animal.

The team was led by OfR Vice President of Universitywide Core Services Jeetendra Eswaraka, Ph.D., and included Rutgers Senior Vice President for Research Michael E. Zwick, Ph.D., and others.

Current methods of checking the state of animals, or assessing animal condition, have limitations. Lab technicians typically conduct assessments during daytime hours, when the animals are less active, and signs of health issues can be subtle and easy to miss.

The innovation the team developed improves the well-being of mice through 24/7 noninvasive tracking of a digital biomarker—locomotor activity—which is then analyzed using a large language model (LLM) to provide alerts to veterinary staff about potential health issues. This AI algorithm detected potential health issues 3–5 days before obvious clinical signs appeared in mice, marking a milestone in health care for mice.

“This project on using machine learning and digital biomarkers to improve the health care of our animals has been a very fulfilling experience,” said Eswaraka.

“The ability to find animals that are likely to get sick 2–3 days before visible symptoms present themselves provides us with an opportunity to really make a difference in improving the health of the animals. Improving veterinary care and improving operational efficiency by over 50% are examples of how AI can change the world around us. I am grateful that Rutgers is developing and adopting these novel technologies, and I thank my collaborators for their hard work and dedication to this development.”

“AI programs such as this one that Eswaraka and our worldwide team developed are changing the way research is being conducted, and the Office for Research is at the forefront, using all available tools and sometimes creating more advanced tools to better support Rutgers researchers,” said Zwick. “I am proud to have been a part of this project, and I look forward to seeing its impact on research around the world.”

The Rutgers University Animal Care Program (RUAC) was recently reaccredited by AAALAC International (AAALACi)—a globally recognized organization that promotes the humane and responsible treatment of animals in science through assessment and accreditation.

The reaccreditation confirms that RUAC not only meets but exceeds the highest ethical, regulatory and technological standards, ensuring continued support for world-class research.

Researchers at the Helmholtz-Zentrum Dresden-Rossendorf (HZDR), together with Wismut GmbH and scientists from the University of Granada in Spain, have demonstrated for the first time that bacteria can convert uranium dissolved in water into a stable chemical compound when they have access to glycerol as a food source.

In the process, uranium assumes a chemical state previously known only as a transient state. The results have been published in the journal Nature Communications and are relevant to future research on the use of bacteria for environmental remediation.

Bacteria in the environment, in soil or water, play an important role in ecosystems. Some specialize in breaking down harmful substances. “There are bacteria that can metabolically utilize the toxic heavy metal uranium,” says Dr. Evelyn Krawczyk-Bärsch, a scientist in HZDR’s Terrestrial Microbiology research group and co-author of the study.

“Our group’s investigations have already revealed that bacteria can use uranium dissolved in water for their metabolism when they have access to glycerol as a food source.” Glycerol is a basic component of plant and animal fats. In nature, for example, it is formed when wood is decomposed by fungi.

But to what extent can bacteria reduce the amount of dissolved uranium in water? And into which chemical forms is free uranium converted by bacterial metabolic processes? These were the questions the researchers addressed in the new study.

Uranium in cell walls

For their experiments, they used mine water from a flooded uranium mine in the Ore Mountains belonging to Wismut GmbH. In laboratory experiments conducted in an oxygen-free environment, the research team added a specific amount of glycerol to the water samples.

“We wanted to create natural conditions for the bacterial community already existing in the mine water because, at a depth of approximately 2,000 meters (6,600 feet), there is usually little or no oxygen in the mine,” explains Antonio M. Newman-Portela, a former doctoral candidate at both HZDR and the Microbiology Department at the University of Granada (Spain), and the lead author of the study.

Under conditions favorable for bacterial growth, the bacteria accepted glycerol as a food source. “After 130 days, only around 5% of the uranium dissolved in the water remained in the samples,” says Newman-Portela.

“We suspected that the bacteria had incorporated the uranium into their cell walls. We already knew about accumulation processes from the literature.” And, indeed, the researchers were able to prove the existence of uranium in the bacteria’s cell walls.

Unusual chemical state

But which precise chemical compounds were involved? To establish this, the team used advanced microscopic and spectroscopic methods. The studies comprised experiments at the Rossendorf Beamline (ROBL), operated by HZDR at the European Synchrotron Radiation Facility (ESRF) in Grenoble, France, as well as complementary studies at the University of Granada.

Initially, the scientists explored the bacterial membrane to establish the chemical states in which the uranium was present. In chemical terminology, the term “valency” is used to describe how many “hands” an atom has to hold onto other atoms within a chemical compound.

“Uranium usually occurs with a valency of 4 or 6. Pentavalent uranium does exist, but it is rare or only transient. Until now, it had been seen in an unstable oxidation state,” explains Newman-Portela. “So, the findings of our study were extremely surprising because in the biomass analyzed from our experimental runs, an unusually high proportion of the uranium identified was also pentavalent uranium.”

Stable—even under the influence of oxygen

Furthermore, the researchers found that the pentavalent uranium formed the compound FeU(V)O₄ with iron and oxygen. “This uranium compound doesn’t yet have a name because it is comparatively new. It was first demonstrated in a 2020 study in which soil samples from parts of Croatia contaminated by uranium ammunition were analyzed,” explains Krawczyk-Bärsch.

“It was found that even under the influence of atmospheric oxygen, this uranium compound had remained stable for more than 25 years. But until now, we didn’t know how this compound is formed in nature or that bacteria play a role in its formation.”

In further experiments, the HZDR research team observed that the amount of FeU(V)O4 actually increased when the dried biomass was exposed to oxygen.

“Our study has revealed for the first time that bacteria supplied with glycerol as a carbon source can convert toxic uranium dissolved in water into a stable chemical compound,” says Krawczyk-Bärsch. “We still have to investigate to what extent bacteria might help render uranium harmless for remediation purposes.”

In future work, the HZDR team aims to gain further insights into uranium-binding bacteria and better understand the underlying biochemical and geochemical processes.

RNA has emerged as one of the most promising molecules in modern medicine, enabling advances from mRNA vaccines and gene therapies to genome editing and synthetic biology. However, designing RNA molecules that reliably fold into a desired secondary structure remains a major challenge. Even for relatively short sequences, the number of possible nucleotide combinations grows exponentially, making it difficult to identify optimal candidates. As a result, conventional computational methods often require extensive candidate evaluations, creating a significant bottleneck when experimental validation is both time-consuming and costly.

To address this challenge, researchers from Keio University, led by Project Lecturer Shuta Kikuchi of the Graduate School of Science and Technology and Professor Shu Tanaka of the Department of Applied Physics and Physico-Informatics, developed a novel RNA inverse folding framework based on factorization machine with quadratic optimization annealing (FMQA). This machine learning– and Ising machine–driven black-box optimization approach is designed to identify high-quality RNA sequence candidates with relatively few evaluations.

“We investigated a new application of FMQA in biomolecular design, where its potential remains relatively unexplored. Since RNA, DNA and protein sequences are inherently categorical in nature, it is unclear how converting them into binary representations affects optimization performance. In this study, we examined RNA inverse folding and the influence of different encoding and assignment choices within FMQA,” says Dr. Kikuchi. The findings are published in Scientific Reports.

The researchers formulated RNA inverse folding as an optimization problem aimed at identifying sequences most likely to fold into a predefined target structure. FMQA served as the core optimization engine, and its performance was evaluated across four binary encoding methods—one-hot, domain-wall, binary and unary—alongside all possible nucleotide-to-integer assignments for adenine (A), uracil (U), guanine (G) and cytosine (C). RNA design quality was assessed using the Normalized Ensemble Defect (NED), which measures the agreement between predicted and target structures. FMQA was benchmarked against random search, genetic algorithms and Bayesian optimization.

The results showed that the encoding strategy plays a decisive role in artificial intelligence and Ising machine–driven RNA design. One-hot and domain-wall encodings consistently outperformed binary and unary representations, producing sequences with lower NED values and higher success rates. Importantly, domain-wall encoding introduced a search bias toward specific integer states. When guanine (G) and cytosine (C) were assigned to these favored states, G–C base pairs accumulated more frequently in stem regions, resulting in greater thermodynamic stability and improved design performance.

Across benchmarks, FMQA also identified high-quality RNA designs with fewer function evaluations than competing methods, demonstrating strong efficiency in search-constrained settings.

Beyond RNA inverse folding, the findings carry broader implications for computational biology and optimization science. They demonstrate that annealing-based optimization frameworks such as FMQA can be effectively extended to life-science problems, strengthening the bridge between quantum-inspired computing and biomolecular engineering. More importantly, the study highlights that data encoding is not merely a preprocessing step but a design variable that can fundamentally shape optimization outcomes. These insights may guide future applications of FMQA in biomolecular design, materials discovery and polymer engineering.

Looking ahead, this approach could accelerate the design of functional biomolecules, particularly RNA systems that must reliably adopt specific structures for therapeutic or diagnostic applications. “Potential applications include biosensors, genome-editing tools, aptamers, ribozymes and riboswitches,” notes Kikuchi. “Because DNA, RNA and proteins are all represented by categorical biological sequences, the approach may also be extended to broader biomolecular design.”

Furthermore, because FMQA is a flexible black-box optimization framework, future implementations could incorporate experimentally measured properties such as molecular stability, binding affinity or gene-expression control, helping to bridge computational design and laboratory validation.

“The insights gained from this study are not limited to RNA,” adds Tanaka. “They have a generality that allows them to be applied to discrete design problems where each evaluation is costly, including materials and molecular design.” In the long term, such evaluation-efficient optimization strategies may help reduce the experimental burden and accelerate discovery across biotechnology and medicine.

“Because FMQA formulates the learned surrogate model as a quadratic optimization problem, it can be implemented with quantum annealing machines,” says Kikuchi. “This perspective points to an exciting future direction: advancing ‘Quantum for Biology’ by exploring how next-generation quantum and quantum-inspired computing technologies can support biomolecular design.”

This study establishes FMQA as a powerful and evaluation-efficient framework for RNA inverse folding. It also highlights a key but often overlooked insight: The way biological sequences are encoded can be as influential as the optimization algorithm itself. Together, these findings open new directions for more efficient, scalable and effective approaches to biomolecular design.

A smartwatch can tell us the level of oxygen in our blood, when our sleep is restless or the number of steps we take in a day. Now imagine that kind of tracking ability for plants. By the time farmers see curling leaves or stunted growth in their fields, their crops may already have spent days under stress.

A new innovation in plant “wearable” sensors aims to catch those distress signals earlier—before the plant visibly suffers, allowing farmers to respond and help their crops thrive.

In a recent study, researchers created tiny tattoo-like sensors that adhere to leaf surfaces and a stretchable band that wraps around stems. Together, they track two vital signs of plant life—the temperature and humidity beneath the leaf’s surface, and whether the stem is still growing. Even more striking, the system runs without an external battery, scavenging power from moisture evaporating from the plant itself.

The work is published in the journal ACS Applied Materials & Interfaces.

“The larger promise is not merely that one plant can wear one sensor,” said Sameer Sonkusale, professor of electrical and computer engineering at Tufts and senior researcher on the project.

“It is that fields could one day contain networks of plant-level monitors, each reporting early signs of thirst, salt stress, disease or nutrient imbalance. Satellites and drones already give farmers a bird’s-eye view. Plant wearables could provide something more intimate: the plant’s-eye view.”

Current methods for monitoring crops use satellite imagery and drones to get visible, infrared and microwave data that map greenness, uneven growth, temperature, pest damage, soil moisture and other big-picture measurements of crop stress. Soil sensors can measure moisture, temperature, pH and some nutrient levels. And weather stations provide information on air temperature, humidity, rainfall, wind and sun exposure.

While those measurements are useful, they focus on conditions that may affect the crops in the future or on an assessment of damage already done. “The leaf sensor is more of an early warning system showing how the plant is responding in the moment, before visible signs appear,” said Nafize Hossain, a graduate student who led the research in the Sonkusale lab.

The sensors can also be extended to track other important indicators of plant health, such as levels of important nutrients and plant hormones that are early signals of root, leaf, stem and fruit growth, as well as responses to pathogens.

Stress trackers

Resembling a temporary tattoo, the leaf sensor is thin, flexible and can sit on uneven surfaces, allowing the plant to breathe and bend in the wind without damage. “Other plant sensors exist, but their ability to track multiple stressors and growth-related parameters is limited, and the technology often relies on external batteries, which complicate field deployment,” said Hossain.

The sensor developed by the researchers provides information on the “vapor pressure deficit,” or VPD. It’s a technical term, but it describes something very intuitive—how likely the air is to pull water from the plant.

When VPD is high, the air is dry and pulls moisture from leaves more aggressively. Plants respond by closing their stomata, the tiny pores that regulate gas exchange and water loss. That can protect them from dehydration, but it also slows photosynthesis and growth.

The leaf moisture sensor uses vanadium pentoxide crystals separated into extremely thin “nanosheets.” The nanosheets are stacked into layers and arranged in a membrane. Another layer of graphene (made of carbon atoms) forms a sieve to let moisture through from the plant to the nanosheets.

When that happens, the water forms ions which sweep through sheets, creating a current—and, voila, it’s not only a sensor but also a battery. The level of the current is directly proportional to the amount of moisture exchange with the air.

The power is tiny—microwatts—but enough, along with low-power electronics and energy storage, to support periodic sensing.

Two signals, different timescales

The stem-based device borrows from kirigami, the Japanese art of cutting paper, so it can stretch and deform in controlled ways. The sensor is coated with a eutectogel, a soft, ion-conducting gel that changes electrical resistance as the stem expands or contracts. In healthy growth, the stem diameter tends to increase. Under stress, growth may slow or the stem may even shrink.

Pairing the two types of sensors is important because plants can show stress on more than one time scale. Leaf sensors, for example, can show whether the plant is facing immediate conditions that drive water loss, while stem growth captures a slower biological process.

In tests on bell pepper plants, the system distinguished healthy plants from plants facing water deficit and salinity stress. Healthy plants showed rhythmic VPD changes over time, following normal daily cycles of air moisture.

Water-stressed plants showed a rising VPD trend. Salinity-stressed plants showed a different pattern, with reduced VPD compared with controls, likely linked to altered water uptake and stomatal behavior. Meanwhile, the stem sensor tracked growth in healthy plants and shrinking or reduced diameter in stressed plants.

Built for field use

The sensors are built with field conditions in mind. The leaf sensor is designed to tolerate bending and stretching, while the stem sensor’s kirigami pattern helps distribute strain and reduces the effects of abrupt disturbances like strong winds.

The team is currently working on a fully functional wireless communication platform for the sensors using LoRa—long range—or Bluetooth-based communication standards.

Farmers are getting more tools in their toolbox, thanks to new research from the University of Missouri that shows how they can tweak planting practices to make the most of every acre.

Traditionally, farmers have applied a one-size-fits-all approach to planting. But that strategy may be leaving profits behind, as not every part of a field performs the same.

With the help of digital technology and artificial intelligence, Mizzou researchers are exploring strategies that allow farmers to manage different parts of a field based on their unique conditions.

“Fields might look the same from the road, but they’re not,” Jasmine Neupane, assistant professor of agricultural systems technology at Mizzou’s College of Agriculture, Food and Natural Resources and co-corresponding author, said. “Some areas have better soil and moisture, while others are more prone to erosion or nutrient loss.”

Smarter planting with AI

In many cases, the strategy of planting more seeds to boost production can increase costs without improving returns. To address this, Neupane and her colleagues used an AI model to analyze data from two Ohio farms and show how variable-rate seeding (VRS) can help farmers strike a better balance.

Instead of planting the same number of seeds across an entire field, VRS allows equipment to adjust seeding rates in real time based on each area’s yield potential. The researchers trained their AI model using common field data—including soil samples, elevation and years of yield records—to generate more location-specific recommendations.

“AI helps farmers choose the right planting rate for different parts of the field,” Neupane said. “It also helps them adjust how much fertilizer and crop protection they use, leading to lower costs and better overall results.”

This targeted approach can also help farmers use resources more efficiently.

“It keeps farmers from applying nutrients or chemicals unnecessarily,” Neupane said. “That helps prevent runoff and other environmental impacts, protecting nearby soil and water.”

Mixed results

Researchers studied corn and soybeans—two of the most common crops in the United States—and found key differences in how they respond to the new approach.

Corn showed consistent and predictable results. The model identified where higher planting rates paid off and where they didn’t, making it a strong candidate for immediate use in precision agriculture.

Soybeans, however, proved to be more complicated because the plants can adapt to the season based on weather conditions. This resilience makes it harder to predict how changes in seeding rates will affect final yields. In many cases, environmental factors such as rainfall and temperature had a greater impact than planting decisions alone, suggesting more research is needed before the model can deliver consistent recommendations for soybean farmers.

This summer, Neupane plans to expand the research to fields at Mizzou’s Digital Agriculture Research and Extension Center. For her, the motivation behind this work is deeply personal.

Growing up in Nepal, Neupane witnessed firsthand the challenges farmers face with small landholdings and limited access to technology. Her goal is to use digital tools and data-driven insights to make farming more accessible and efficient for those growing food around the world.

As AI tools continue to advance, she hopes these technologies will help farmers better understand their fields and make more informed decisions.

“When you really understand what your field is telling you, you can manage it much more strategically,” Neupane said.

The study, “Leveraging machine learning and geospatial analysis to determine agronomic and economic optima for variable-rate seeding in corn and soybean,” was published in the Agronomy Journal.

The speed at which a cell produces proteins is a decisive factor in determining whether it divides, specializes or retains its stem cell properties. A team of researchers led by Professor Stefan H. Stricker, professor of epigenetic engineering at LMU’s Biomedical Center and research group leader at Helmholtz Munich, has worked with international partners to demonstrate directly for the first time that the amount of ribosomal RNA (rRNA) directly regulates these processes. Their results were published in the journal Science.

New method makes it possible to control ribosomal RNA in a targeted manner

It has been established for some time that the amount of ribosomal RNA differs among different types of cells and is altered in a number of diseases. But it remained unclear whether these specific characteristics are the cause or merely the result of biological processes.

With the newly developed CRISPR-based method TAPIR (Targeted Activation of Protein Translation), researchers now have access to a tool that can boost the activity of ribosomal genes and, as a result, influence a cell’s protein production. “Our new study shows that targeted activation of rRNA production significantly increases protein synthesis,” explains Stricker, lead author of the publication.

New perspectives for rare diseases and cancer

The results could be particularly relevant for diseases in which ribosome function is disrupted. These include ribosomopathies such as Treacher-Collins syndrome, a rare congenital disease that causes facial malformations. In a mouse model, the researchers partially compensated for disease-related alterations by stimulating rRNA production in a targeted way.

In addition, the research team observed that similar mechanisms also play a role in pancreatic cancer. Tumor cells seem to use increased rRNA production to maintain their rapid growth. In the mouse model for pancreatic cancer, TAPIR was able to increase rRNA production and promote the growth of the cancer cells. This shows that increased rRNA production has a causal effect in contributing to tumor growth and is not just a side effect.

A platform for further research topics related to health

“Our study clearly shows that the regulation of protein biosynthesis plays a key role both in processes of development and growth and in the development of cancer,” Stricker says in summary. He views TAPIR as a research platform for better understanding the impact of protein synthesis on health and disease, and for developing new therapeutic approaches over the long term.

It is conceivable that in the future this approach will be suitable for treating diseases associated with reduced ribosome function and also open up new targets for therapies to combat tumors in which protein production has spiraled out of control.