AI software revolutionizes plant engineering to combat climate change

At Salk in La Jolla, researchers are collaborating to harness the power of artificial intelligence (AI) to engineer plants that can help combat climate change. They are using a pioneering deep learning software called SLEAP to optimize plant root systems, which will capture and store carbon dioxide from the atmosphere. This innovative research offers a promising solution to mitigate the impacts of global warming.

Originally developed for tracking animal movement, SLEAP has been repurposed by Salk Fellow Talmo Pereira and plant scientist Professor Wolfgang Busch to analyze plant root growth with extraordinary precision and efficiency. The scientists have unlocked a sophisticated tool to expedite the design of climate-saving plants by utilizing this state-of-the-art AI software. This is a pivotal endeavor advocated by the Intergovernmental Panel on Climate Change (IPCC) to limit global temperature rise. From left: Talmo Pereira, Elizabeth Berrigan, and Wolfgang Busch.

The study published in Plant Phenomics on April 12, 2024, introduced a new protocol for utilizing SLEAP to analyze various aspects of plant root systems. These include depth, mass, and angle of growth. The painstaking process of manually measuring these physical characteristics posed significant challenges and time constraints to researchers before the advent of SLEAP. The innovative combination of computer vision and deep learning incorporated within SLEAP has revolutionized this paradigm, enabling researchers to achieve accurate and rapid analysis of plant root features without the cumbersome, frame-by-frame manual labor required by previous AI models.

One of the most striking achievements resulting from the application of SLEAP to plants is the development of the most extensive catalog of plant root system phenotypes to date. This invaluable resource facilitates the identification of genes associated with specific root characteristics and elucidates the complex relationships between different root traits, providing crucial insights into the genes most beneficial for optimizing plant designs.

The transformative capabilities of SLEAP were further demonstratedthrough the creation of the sleap-roots toolkit, open-source software that empowers SLEAP to process biological traits of root systems. This toolkit not only expedited the analysis of plant images but also outperformed existing practices by annotating 1.5 times faster, training the AI model 10 times faster, and predicting plant structure on new data 10 times faster, all while maintaining or even improving accuracy.

By seamlessly connecting phenotype and genotype data, SLEAP and the sleap-roots toolkit are poised to revolutionize the efforts of Salk's Harnessing Plants Initiative to engineer plants with enhanced carbon-capturing capabilities and deeper, more robust root systems. These advancements hold the potential to accelerate the development of climate-resilient plants that can significantly mitigate the impacts of climate change.

SLEAP has not only positioned Salk as a trailblazer in plant engineering but has also garnered attention from scientists at NASA, reflecting the global impact and potential of this pioneering technology. With accessibility and reproducibility at the forefront of its design, SLEAP and the sleap-roots toolkit offer an invaluable resource to researchers worldwide, heralding a new era of plant engineering and environmental conservation.

As the collaborative team at Salk embarks on new frontiers, including the analysis of 3D data using SLEAP, the profound impact of this deep learning software on accelerating plant designs and shaping the future of climate change research is already palpable. The journey to refine, expand, and share SLEAP and the sleap-roots toolkit is poised to continue for years to come, cementing their vital role in advancing scientific endeavors and making a significant impact in the global fight against climate change.

This exploration of SLEAP's potential to engineer plants reflects a convergence of diverse disciplines, showcasing the remarkable potential of AI-led innovation in shaping a sustainable future and fostering interdisciplinary collaboration to create profound and transformative scientific advancements.

Data detecting a burst was sent to the Integral Science Data Center in Geneva. The software analyzed the data and determined that the burst originated from the nearby M82 galaxy. The burst's location is marked with a small square on Integral's map, and the corresponding location is indicated by a blue circle on the two cut-out images. This information was made possible through the combined efforts of ESA/Integral, ESA/XMM-Newton, INAF/TNG, and M. Rigoselli (INAF).
Data detecting a burst was sent to the Integral Science Data Center in Geneva. The software analyzed the data and determined that the burst originated from the nearby M82 galaxy. The burst's location is marked with a small square on Integral's map, and the corresponding location is indicated by a blue circle on the two cut-out images. This information was made possible through the combined efforts of ESA/Integral, ESA/XMM-Newton, INAF/TNG, and M. Rigoselli (INAF).

University of Geneva researches a massive magnetic star eruption that is illuminating a nearby galaxy

A groundbreaking discovery has been made by the University of Geneva in a recent publication concerning the eruption of a mega-magnetic star that illuminated a neighboring galaxy. An international team of researchers, including UNIGE researchers, identified a rare cosmic phenomenon involving an extremely magnetic neutron star,known as a magnetar. The observational data was collected by the European Space Agency's (ESA) satellite, INTEGRAL, which detected a burst of gamma rays coming from the nearby galaxy M82. After the detection, the ESA's XMM-Newton X-ray space telescope was used to search for any afterglow from the explosion, but it yielded no results.

The automatic data processing system, particularly the IBAS (Integral Burst Alert System), provided a crucial automatic localization of the event coinciding with the galaxy M82, situated 12 million light-years away. The IBAS was developed and operated by scientific and engineering teams from the University of Geneva in collaboration with international partners. "One of the most striking aspects of this discovery is the rapid alert dissemination enabled by our automatic data processing system," stated Carlo Ferrigno, senior research associate at UNIGE's Astronomy Department. The system's ability to promptly localize such significant cosmic events is paramount in facilitating timely follow-up observations, as emphasized by the immediate request for XMM-Newton's follow-up observation post the gamma-ray burst detection.

This discovery signifies a milestone in the elucidation of these enigmatic cosmic phenomena as it marks the first confirmed instance of a magnetar flare outside our galaxy. "The search for additional magnetars in other extra-galactic star-forming regions is crucial to comprehending the frequency and mechanisms behind these flare events," added Volodymyr Savchenko, another senior research associate at UNIGE's Faculty of Science.

The INTEGRAL satellite played an invaluable role in this discovery, coupled with the sophisticated automatic data processing system, demonstrating the pioneering nature of the research conducted by the University of Geneva. As the inclusive collaboration of international researchers continues to unveil the mysteries of the universe, the inherent significance of advanced data processing systems in facilitating such discoveries remains undeniable. According to the researchers, this recent cosmic event sheds new light on the understanding of magnetars and neutron stars, providing crucial insights into the mechanisms governing these highly magnetic and energetic celestial bodies.

Supercomputer simulations uncover the dynamics of undercurrent accelerating the melting of ice shelves

New research conducted by a team of scientists from the University of Southampton, UK has provided new insights into the melting of floating sections of the West Antarctic Ice Sheet and how it contributes to rising global sea levels. The study, published in the prestigious journal Science Advances, sheds light on the mechanisms that drive the melting of ice shelves beneath the ocean's surface.

The West Antarctic Ice Sheet has been losing significant mass in recent decades, which has contributed to the global rise in sea levels. If the entire ice sheet melts, it could lead to a five-meter increase in sea levels worldwide. The study highlights the role played by the Circumpolar Deep Water (CDW) which is up to 4°C warmer than the local freezing temperatures and flows beneath the ice shelves in West Antarctica, causing them to melt from below. Since a significant portion of the West Antarctic Ice Sheet is below sea level, it is particularly vulnerable to the intrusion of this warm water, which could cause the ice to retreat further in the future.

The researchers used high-resolution simulations on supercomputers to investigate the complex dynamics of the undercurrent that transports the warm water towards the ice shelves. They discovered that when the warm CDW interacts with the ice shelf, it not only melts the ice but also mixes with the lighter, melted freshwater. As this mixed water rises through the water layers above, it spreads out and stretches the layer of CDW vertically, creating a swirling motion in the water. If a trough is present near the coast, this swirling motion carries the water away from the ice shelf cavity and toward the shelf's edge due to the movement of pressure within the fluid. This movement generates a current along the seafloor slope, directing more warm water towards the ice shelf. The underwater current originates slightly farther away from the ice shelf and intensifies as more ice melts, transporting even greater amounts of warm water toward the ice shelves.

Dr. Alessandro Silvano from the University of Southampton, coauthor of the study, emphasizes the significance of their findings, stating that "Scientific models that don't include the cavities under ice shelves are probably overlooking this positive feedback loop. Our results suggest it's an important factor that could affect how quickly ice shelves melt and how stable the West Antarctic Ice Sheet is over time."

The study highlights the critical role played by supercomputer simulations in advancing our understanding of climate-related phenomena and their implications for our planet's future. The research conducted by the international collaboration was made possible by support from the National Science Foundation and the Natural Environment Research Council. These groundbreaking findings offer invaluable insights into the current state of Antarctica's ice shelves and their susceptibility to further melting.

This is a false-color scanning electron micrograph that shows lung cancer cells grown in culture. The image is courtesy of Anne Weston. A new AI tool named PERCEPTION is helping predict patients' response to various therapies by utilizing data at the level of individual cells.
This is a false-color scanning electron micrograph that shows lung cancer cells grown in culture. The image is courtesy of Anne Weston. A new AI tool named PERCEPTION is helping predict patients' response to various therapies by utilizing data at the level of individual cells.

AI tool predicts responses to cancer therapy using information from each cell of the tumor

A recent study conducted by researchers from Sanford Burnham Prebys in La Jolla, CA and the National Cancer Institute has highlighted the extraordinary capabilities of a new artificial intelligence (AI) tool called PERsonalized Single-Cell Expression-Based Planning for Treatments in Oncology (PERCEPTION). The tool utilizes machine learning algorithms to analyze information from every cell of a tumor, allowing clinicians to predict how patients will respond to cancer therapy.

Traditional methods of precision oncology treatments have focused on identifying genetic mutations in cancer driver genes and matching patients with targeted therapies accordingly. However, many cancer patients do not benefit from these early targeted therapies. To address this challenge, the team of researchers led by Dr. Sanju Sinha developed a new computational pipeline to predict patient response to cancer drugs at the single-cell level.

The team used transcriptomics, the study of transcription factors, to develop PERCEPTION. The tool provides a deep understanding of the clonal architecture of the tumor and can even detect the emergence of drug resistance by analyzing messenger RNA molecules expressed by genes. This ability to monitor resistance offers the potential for treatment modification and adaptation to the evolving nature of cancer cells.

To build PERCEPTION, the researchers employed transfer learning, a branch of AI, to leverage limited single-cell data from clinics. The tool was pre-trained using published bulk-gene expression data from tumors and then fine-tuned using single-cell data from cell lines and patients. The researchers successfully validated PERCEPTION by accurately predicting patient responses to monotherapy and combination treatments in three independently conducted clinical trials for multiple myeloma, breast, and lung cancer.

Dr. Sanju Sinha cautions that while PERCEPTION holds tremendous promise, it is not yet ready for clinical use. However, its success in predicting treatment responses underscores the potential of using single-cell information to guide personalized treatment strategies. The researchers hope that these findings will encourage the adoption of PERCEPTION in clinics, generating more data that can be used to refine and further develop the tool for clinical use. Sanju Sinha

"The quality of the prediction rises with the quality and quantity of the data serving as its foundation," says Dr. Sinha. "Our goal is to create a clinical tool that can predict the treatment response of individual cancer patients in a systematic, data-driven manner. We hope these findings spur more data and more such studies, sooner rather than later."

The development of PERCEPTION represents a significant step forward in the field of precision oncology, offering hope for improved treatment outcomes and enhanced patient care. As the researchers continue to refine this AI tool, its potential impact on cancer therapy is monumental. The groundbreaking research was supported by funding from the Intramural Research Program of the National Institutes of Health (NIH), the National Cancer Institute (NCI), and various NIH grants.

The Near Space Network developed new antennas in Alaska, Chile, Norway, and Virginia in partnership with KSAT.
The Near Space Network developed new antennas in Alaska, Chile, Norway, and Virginia in partnership with KSAT.

NASA's Near Space Network enables the PACE Climate Mission to establish communication with Earth

NASA's PACE mission achieved a significant milestone by successfully transmitting its first operational data back to scientists and researchers. This was made possible, in part, by NASA's Near Space Network's innovative data-storing technology, which introduced two key enhancements for PACE and other upcoming science missions. 

When a satellite orbits in space, it generates crucial data about its health, location, battery life, and more. At the same time, the mission's scientific instruments capture images and data that support the overall objective of the satellite. However, transmitting this data back to Earth poses several challenges, which include extreme distances and disruptions or delays that can occur during transmission.

To tackle these challenges, NASA's Near Space Network integrated Delay/Disruption Tolerant Networking (DTN) into four new antennas and the PACE spacecraft. DTN allows for the safe storage and forwarding of data when disruptions occur, ensuring that important information is not lost.

Kevin Coggins, Deputy Associate Administrator for NASA's Space Communications and Navigation (SCaN) program, stressed the importance of DTN, stating, "DTN is the future of space communications, providing robust protection of data that could be lost due to a disruption. PACE is the first operational science mission to leverage DTN, and we are using it to transmit data to mission operators monitoring the batteries, orbit, and more. This information is critical to mission operations."

The PACE mission, located approximately 250 miles above Earth, aims to collect data that helps researchers better understand carbon dioxide exchange between the ocean and atmosphere, monitor air quality and climate-related atmospheric variables, and study the health of the ocean by examining phytoplankton.

While PACE is the first operational science user of DTN, demonstrations of the technology have been successfully conducted on the International Space Station. In addition to DTN, the Near Space Network collaborated with commercial partner Kongsberg Satellite Services in Norway to integrate four new antennas into the network.

These antennas, located in Fairbanks, Alaska; Wallops Island, Virginia; Punta Arenas, Chile; and Svalbard, Norway, allow missions to downlink terabytes of science data at once. As PACE orbits Earth, it will downlink its science data 12 to 15 times a day to three of the network's new antennas, resulting in a daily transmission of 3.5 terabytes of science data.

These advancements in network capability, including DTN and the new antennas, contribute to the Near Space Network's mission to support science missions, human spaceflight, and technology experiments.

Deputy Associate Administrator Kevin Coggins expressed his satisfaction with NASA's Near Space Network, stating, "NASA's Near Space Network now has unprecedented flexibility to get scientists and operations managers more of the precious information they need to ensure their mission's success."

In addition to these new capabilities, the network is also expanding its portfolio by increasing the number of commercial antennas. In 2023, NASA issued a request for proposal seeking commercial providers to integrate into the growing portfolio of the Near Space Network. With an enhanced capacity, the network can support additional science missions and provide more opportunities for data transmission.

The Near Space Network, funded by NASA's Space Communications and Navigation (SCaN) program office at NASA Headquarters in Washington, operates from NASA's Goddard Space Flight Center in Greenbelt, Maryland, and these recent enhancements mark significant progress in advancing communication systems for missions near Earth and in deep space.