Recent years have seen an explosion in profiling single cells from a diverse range of human tissue and organs — such as a neurons, muscles, and immune cells — to gain insight into human health and treating disease. The largest datasets contain anywhere from around 100,000 to 2 million cells, and growing. The long-term goal of the Human Cell Atlas, for instance, is to profile about 10 billion cells. Each cell itself contains tons of data on RNA expression, which can provide insight about cell behavior and disease progression.

With enough computation power, biologists can analyze full datasets, but it takes hours or days. Without those resources, it’s impractical. Sampling methods can be used to extract small subsets of the cells for faster, more efficient analysis, but they don’t scale well to large datasets and often miss less abundant cell types.

In a paper being presented next week at the Research in Computational Molecular Biology conference, the MIT researchers describe a method that captures a fully comprehensive “sketch” of an entire dataset that can be shared and merged easily with other datasets. Instead of sampling cells with equal probability, it evenly samples cells from across the diverse cell types present in the dataset.

“These are like sketches on paper, where an artist will try to preserve all the important features of a main image,” says Bonnie Berger, the Simons Professor of Mathematics at MIT, a professor of electrical engineering and computer science, and head of the Computation and Biology group.

In experiments, the method generated sketches from datasets of millions of cells in a few minutes — as opposed to a few hours — that had far more equal representation of rare cells from across the datasets. The sketches even captured, in one instance, a rare subset of inflammatory macrophages that other methods missed.

“Most biologists analyzing single-cell data are just working on their laptops,” says Brian Hie, a PhD student in the Computer Science and Artificial Intelligence Laboratory (CSAIL) and a researcher in the Computation and Biology group. “Sketching gives a compact summary of a very large dataset that tries to preserve as much biological information as possible … so people don’t need to use so much computational power.”

Joining Hie and Berger on the paper are: CSAIL PhD student Hyunghoon Cho; Benjamin DeMeo, a graduate student at MIT and Harvard Medical School; and Bryan Bryson, an MIT assistant professor of biological engineering.

Plaid coverings

Humans have hundreds of categories and subcategories of cells, and each cell expresses a diverse set of genes. Techniques such as RNA sequencing capture all cell information in massive tables, where each row represents a cell and each column represents some measurement of gene expression. Cells are points scattered around a sprawling multidimensional space where each dimension corresponds to the expression of a different gene.

As it happens, cell types with similar gene diversity — both common and rare — form similar-sized clusters that take up roughly the same space. But the density of cells within those clusters varies greatly: 1,000 cells may reside in a common cluster, while the equally diverse rare cluster will contain 10 cells. That’s a problem for traditional sampling methods that extract a target-size sample of single cells.

“If you take a 10-percent sample, and there are 10 cells in a rare cluster and 1,000 cells in a common cluster, you’re more likely to grab tons of common cells, but miss all rare cells,” Hie says. “But rare cells can lead to important biological discoveries.”

The researchers modified a class of algorithm that lays shapes over datasets. Their algorithm covers the entire computational space with what they call a “plaid covering,” which is like a grid of equal-sized squares but in many dimensions. It only lays these multidimensional squares where there’s at least one cell, and skips over any empty regions. In the end, the grid’s empty columns will be much wider or skinnier than occupied columns — hence the “plaid” description. That technique saves tons of computation to help the covering scale to massive datasets.

Capturing rare cells

Occupied squares may contain only one cell or 1,000 cells, but they will all have the exact same sampling weight. The algorithm then finds a target sample — of, say, 20,000 cells — by selecting a set number of cells from each occupied square uniformly, at random. The resulting sketch contains a far more equal distribution of cell types — for example, 10 common cells from a cluster of 100 and eight rare cells from a cluster of 10.

“We take advantage of these cell types occupying similar volumes of space,” Hie says. “Because we sample according to volume, instead of density, we get a more even coverage of the biological space … and we’re naturally preserving the rare cell types.”

They applied their sketching method to a dataset of around 250,000 umbilical cord cells that contained two subsets of a rare macrophages — inflammatory and anti-inflammatory. All other traditional sampling methods clustered both subsets together, while the sketching method separated them. Additional in-depth studies of these macrophage subpopulations could help reveal insight into inflammation and how to modulate inflammatory processes in response to disease, the researchers say.

“That’s a benefit in working at the interface of fields,” Berger says. “We’re trained as mathematicians, but we understand what biological data science problems are, so we can bring the best technologies to their analysis.”

Caroline Ng’weno, who conducted the research during her UW graduate studies, is one of the first — if not the first — Kenyan women to have earned a Ph.D. working as a field biologist in Kenya. Ng’weno now heads the Pride of Meru program for the Born Free Foundation, an organization dedicated to research and protection of lions in central Kenya. Her work has provided new insights into the interaction among Jackson’s hartebeest, a species of conservation concern; other wild ungulates; cattle; and lions in Kenya’s Ol Pejeta Conservancy, which is managed for both wildlife conservation and cattle ranching.

As detailed in the scientific journals Ecology and Frontiers in Ecology and Evolution, Jackson’s hartebeest — a large antelope that can weigh over 400 pounds — has been in significant decline in that part of Kenya since the reintroduction of lions in the late 1980s. That’s primarily because hartebeest share savanna habitat with zebras, the primary prey of the approximately 70 lions comprising five prides in the conservancy.

“Predation risk for hartebeest was elevated in association with zebra, implying that apparent competition with zebra may negatively impact hartebeest populations,” Ng’weno and fellow researchers wrote in Ecology. That’s in line with previous research that shows secondary prey such as hartebeest can suffer significant population declines when large carnivores are restored to an ecosystem after a long absence.

Restoration of lions to the Ol Pejeta Conservancy and elsewhere in central Kenya resulted from greater tolerance by ranch managers, following decades of shooting and poisoning of the top predators. That’s because ranch managers are increasingly recognizing that tourism resulting from abundant wildlife populations can help them sustain their livestock operations in drought years and other lean times. However, the decline of hartebeest numbers has led some ranch managers to considering reimplementing lethal control of lions.

According to Ng’weno’s research, the ranchers could hold the key to maintaining both the lion and hartebeest populations.

It turns out that abandoned cattle corrals create nutrient hotspots called glades that attract zebras, and therefore lions — but not hartebeest. And the researchers showed that survival of hartebeest in areas without glades was more than twice as high as in areas with glades.

So, ranch managers’ placement of cattle corrals away from hartebeest likely would allow the antelope species to increase, with lions focused on the zebras that congregate at the resulting glades.

“In our study system, spatial separation between zebra and hartebeest improved survival rates of hartebeest, probably by reducing encounters with lions hunting in areas with high zebra densities,” they wrote. “Strategic placement of glades, therefore, offers a promising approach to creating refuges for hartebeest and perhaps other species of secondary prey.”

While some might argue that eliminating glades through reduction of cattle production might be an option for hartebeest conservation, the researchers say that’s not practical. That’s because ranchers are unlikely to reduce cattle numbers voluntarily; and reducing cattle numbers would likely boost zebra numbers, along with lions, as cattle and zebra diets overlap. The researchers note that predation by lions on cattle is rare compared to predation on zebras.

“Alternative conservation interventions are required for the long-term persistence of lions and their prey not only on Ol Pejeta Conservancy, but more widely in (Kenya’s) Laikipia County and the whole of sub-Saharan Africa,” they wrote.

The research involved capturing and placing GPS collars on lions in five different prides representing 70 individuals; identification and tracking of 179 hartebeest; and analysis of 246 sites where lions killed animals. The scientists also studied lion predation on buffalo, zebras, impala and warthogs. Notably, they found that the proximity of buffalo to zebras reduced the risk of predation on buffalo, contrary to the relative risk for hartebeest.

Story Source:

Materials provided by University of Wyoming. Note: Content may be edited for style and length.

As the population grows and climate change progresses, food resources could become scarce in future. In view of the impending scenarios, plant breeders are faced with the challenge of improving the yield of crop plants. Can existing varieties be optimized through breeding? To help boost the yield of today’s sorts, an international team of scientists studied the genetic diversity of various wheat varieties and in doing so have discovered astonishing relationships with human sociocultural history.

Scientists at Helmholtz Zentrum München were involved in the large-scale WHEALBI study, funded by the European Union. Along with teams from France, Italy, Hungary, Turkey and other European countries, they analyzed the genomes of 480 wheat varieties, including wild grasses, ancient grains and modern high-performance types. In addition to learn about the evolution and cultivation of today’s bread wheat, the geneticists also linked the development of wheat to geographic and geopolitical events in human history.

Modern bread wheat originated around 10,000 years ago in the region of modern-day Turkey from a cross between durum wheat and a wild grass (Aegilops tauschii), while the grain we call spelt stems from cultivated emmer and various types of bread wheat. “The occurrence of cultivated plants is closely linked to human migrations over the millennia,” says bioinformatician Michael Seidel, along with Daniel Lang one of the lead authors of the study. Both researchers work in the Plant Genome and Systems Biology group (PGSB) at the Helmholtz Zentrum München.

The PGSB team identified three gene pools in the bread-wheat varieties used today that are closely linked to historical events: one from high-yielding varieties domesticated in the near east that spread as part of the green revolution and two separate gene pools from Western and Central Europe. They diverged between 1966 and 1985 as a result of geopolitical and socio-economic separation during the Cold War. With the fall of the Iron Curtain in 1989, the wheat lines gradually admixed again, as their genomes reveal.

Even the emergence and expansion of the European Union can be seen in the genome of today’s wheat. Wheat lines that used to be cultivated mainly in Central Europe are now used throughout Europe. “These examples demonstrate the influence of humans on the distribution and evolution of crop plants — beyond their actual development into cultivated plants,” as bioinformatician Lang of Helmholtz Zentrum München stated.

Knowledge of the genetic diversity of wheat is a prerequisite for optimizing modern wheat varieties. Familiarity with the key characteristics for breeding is the essential precondition for rendering future varieties more productive and meeting the demands of a growing world population and imminent climate change. Together with corn and rice, wheat ranks as one of the world’s three most important staple foods. Growing wheat in spite of dwindling soil and water resources in potentially challenging climatic conditions could become vital in the future.

Consequently, the researchers involved in the WHEALBI study identified previously unknown genes that influence the yield, flowering time, height and stability of wheat plants. For corresponding author Georg Haberer of the PGSB this is just the beginning: “We expect a large number of further studies that will make good use of these findings for breeding research.”

The international research team bombarded graphene with hydrogen atoms. “The hydrogen atom behaved quite differently than we expected,” says Alec Wodtke, head of the Department of Dynamics at Surfaces at the Max Planck Institute (MPI) for Biophysical Chemistry and professor at the Institute of Physical Chemistry at the University of Göttingen. “Instead of immediately flying away, the hydrogen atoms ‘stick’ briefly to the carbon atoms and then bounce off the surface. They form a transient chemical bond,” Wodtke reports. And something else surprised the scientists: The hydrogen atoms have a lot of energy before they hit the graphene, but not much left when they fly away. Hydrogen atoms lose most of their energy on collision, but where does it go?

To explain these surprising experimental observations, the Göttingen MPI researcher Alexander Kandratsenka, in cooperation with colleagues at the California Institute of Technology, developed theoretical methods, which they simulated on the computer and then compared to their experiments. With these theoretical simulations, which agree well with the experimental observations, the researchers were able to reproduce the ultra-fast movements of atoms forming the transient chemical bond. “This bond lasts for only about ten femtoseconds — ten quadrillionths of a second. This makes it one of the fastest chemical reactions ever observed directly,” Kandratsenka explains.

“During these ten femtoseconds, the hydrogen atom can transfer almost all its energy to the carbon atoms of the graphene and it triggers a sound wave that propagates outward from the point of the hydrogen atom impact over the graphene surface, much like a stone that falls into water and triggers a wave,” says Kandratsenka. The sound wave contributes to the fact that the hydrogen atom can bind more easily to the carbon atom than the scientists had expected and previous models had predicted.

The results of the research team provide fundamentally new insights into chemical bonding. In addition, they are of great interest to industry. Sticking Hydrogen atoms to graphene can produce a bandgap, making it a useful semiconductor and much more versatile in electronics.

The effort involved in setting up and running these experiments was enormous, revealed Oliver Bünermann, project group leader at the University of Göttingen. “We had to carry them out in ultra-high vacuum to keep the graphene surface perfectly clean.” The scientists also had to use a large number of laser systems to prepare the hydrogen atoms before the experiment and to detect them after the collision. According to Bünermann, the excellent technical staff in the workshops at the MPI for Biophysical Chemistry and at the University of Göttingen were essential to the project’s success.

Story Source:

Materials provided by University of Göttingen. Note: Content may be edited for style and length.

New research for the first time quantifies, through experiments and mathematical modelling, just how large and how dense a continuous meadow of seagrass must be to provide adequate damping of waves in a given geographic, climatic, and oceanographic setting.

In a pair of papers appearing in the May issues of two research journals, Coastal Engineering and the Journal of Fluids and Structures, MIT professor of civil and environmental engineering Heidi Nepf and doctoral student Jiarui Lei describe their findings and the significant environmental benefits seagrass offers. These include not only preventing beach erosion and protecting seawalls and other structures, but also improving water quality and sequestering carbon to help limit future climate change.

Those services, coupled with better-known services such as providing habitat for fish and food for other marine creatures, mean that submerged aquatic vegetation including seagrass provides an overall value of more than $4 trillion globally every year, as earlier studies have shown. Yet today, some important seagrass areas such as the Chesapeake Bay are down to about half of their historic seagrass coverage (having rebounded from a low of just 2 percent), thus limiting the availability of these valuable services.

Nepf and Lei recreated artificial versions of seagrass, assembled from materials of different stiffness to reproduce the long, flexible blades and much stiffer bases that are typical of seagrass plants such as Zostera marina, also known as common eelgrass. They set up a meadow-like collection of these artificial plants in a 79-foot-long (24-meter) wave tank in MIT’s Parsons Laboratory, which can mimic conditions of natural waves and currents. They subjected the meadow to a variety of conditions, including still water, strong currents, and wave-like sloshing back and forth. Their results validated predictions made earlier using a computerized model of individual plants.

The researchers used the physical and numerical models to analyze how the seagrass and waves interact under a variety of conditions of plant density, blade lengths, and water motions. The study describes how the motion of the plants changes with blade stiffness, wave period, and wave amplitude, providing a more precise prediction of wave damping over seagrass meadows. While other research has modeled some of these conditions, the new work more faithfully reproduces real-world conditions and provides a more realistic platform for testing ideas about seagrass restoration or ways of optimizing the beneficial effects of such submerged meadows, they say.

To test the validity of the model, the team then did a comparison of the predicted effects of seagrass on waves, looking at one specific seagrass meadow off the shore of the Spanish island of Mallorca, in the Mediterranean Sea, which is known to attenuate the force of incoming waves by a factor of about 50 percent on average. Using measurements of meadow morphology and wave velocities collected in a previous study led by Professor Eduardo Infantes, currently of Gothenburg University, Lei was able to confirm the predictions made by the model, which analyzed the way the tips of the grass blades and particles suspended in the water both tend to follow circular paths as waves go by, forming circles of motion known as orbitals.

The observations there matched the predictions very well, Lei says, showing the way wave strength and seagrass motion varied with distance from the edge of the meadow to its interior agreed with the model. So, “with this model the engineers and practitioners can assess different scenarios for seagrass restoration projects, which is a big deal right now,” he says That could make a significant difference, he says, because now some restoration projects are considered too expensive to undertake, whereas a better analysis could show that a smaller area, less expensive to restore, might be capable of providing the desired level of protection. In other areas, the analysis might show that a project is not worth doing at all, because the characteristics of the local waves or currents would limit the grasses’ effectiveness.

The particular seagrass meadow in Mallorca that they studied is known to be very dense and uniform, so one future project is to extend the comparison to other seagrass areas, including those that are patchier or less thickly vegetated, Nepf says, to demonstrate that the model can indeed be useful under a variety of conditions.

By attenuating the waves and thus providing protection against erosion, the seagrass can trap fine sediment on the seabed. This can significantly reduce or prevent runaway growth of algae fed by the nutrients associated with the fine sediment, which in turn causes a depletion of oxygen that can kill off much of the marine life, a process called eutrophication.

Seagrass also has significant potential for sequestering carbon, both through its own biomass and by filtering out fine organic material from the surrounding water, according to Nepf, and this is a focus of her and Lei’s ongoing research. An acre of seagrass can store about three times as much carbon as an acre of rainforest, and Lei says preliminary calculations suggest that globally, seagrass meadows are responsible for more than 10 percent of carbon buried in the ocean, even though they occupy just 0.2 percent of the area.

The work was funded by the U.S. National Science Foundation.