Aurelia Butler

Combining machine learning with multimodal electrochemical sensing can significantly improve the analytical performance of biosensors, according to new findings from a Penn State research team. These improvements may benefit noninvasive health monitoring, such as testing that involves saliva or sweat. The findings were published this month in Analytica Chimica Acta.
The researchers developed a novel analytical platform that enabled them to selectively measure multiple biomolecules using a single sensor, saving space and reducing complexity as compared to the usual route of using multi-sensor systems. In particular, they showed that their sensor can simultaneously detect small quantities of uric acid and tyrosine — two important biomarkers associated with kidney and cardiovascular diseases, diabetes, metabolic disorders, and neuropsychiatric and eating disorders — in sweat and saliva, making the developed method suitable for personalized health monitoring and intervention.
Many biomarkers have similar molecular structures or overlapping electrochemical signatures, making it difficult to detect them simultaneously. Leveraging machine learning for measuring multiple biomarkers can improve the accuracy and reliability of diagnostics and as a result improve patient outcomes, according to the researchers. Further, sensing using the same device saves resources and biological sample volumes needed for tests, which is critical with clinical samples with scarce amounts.
“We developed a new approach to improve the performance of electrochemical biosensors by combining machine learning with multimodal measurement,” said Aida Ebrahimi, Thomas and Sheila Roell Early Career Assistant Professor of Electrical Engineering and assistant professor of biomedical engineering. “Using our optimized machine learning architecture, we could detect biomolecules in amounts 100 times lower than what conventional sensing methods can do.”
The researchers’ methodology features a hardware/software system that enables them to automatically gather and process information based on a machine learning model that is trained to identify biomolecules in biological fluids such as saliva and sweat, which are common choices for noninvasive health monitoring.
“The machine learning-powered electrochemical diagnostic approach presented in this paper may find broader application in multiplexed biochemical sensing,” said Vinay Kammarchedu, 2022-23 Milton and Albertha Langdon Memorial Graduate Fellow in Electrical Engineering at Penn State and first author on the paper. “For example, this method can be extended to a variety of other molecules, including food and water toxins, drugs and neurochemicals that are challenging to detect simultaneously using conventional electrochemical methods.”
In their ongoing work, the researchers are applying this approach on such neurochemicals, which are difficult to detect due to similarities in their molecular structure and overlapping electrochemical signatures. More

Researchers at The University of Texas at Dallas and Novartis Pharmaceuticals Corp. have devised a computer-based platform for drug discovery that could make the process more effective, more efficient and less costly.
Dr. Baris Coskunuzer, professor of mathematical sciences at UT Dallas, and his colleagues developed an approach based on topological data analysis to screen thousands of possible drug candidates virtually and narrow the compound candidates considerably to those that are most fit for laboratory and clinical testing.
The researchers will present their findings at the 36th Conference on Neural Information Processing Systems, which will be held Nov. 28 through Dec. 9 in New Orleans.
Typically, the early phases of drug discovery involve researchers identifying a biological target, such as a protein associated with a disease of interest. The next step is to screen libraries of thousands of potential chemical compounds that might be effective or could be modified to affect the target to alleviate the disease’s cause or symptoms. The most promising candidates move on to the lengthy and expensive process of laboratory and clinical testing and regulatory approval.
“The drug-discovery process can take 10 to 15 years and cost a billion dollars,” Coskunuzer said. “Drug companies want a more cost-effective way to do this. They want to find the most promising compounds at the beginning of the process so they’re not wasting time testing dead ends.
“We have provided a completely new method of virtual screening that is computationally efficient and ranks compounds based on how likely they are to work.”
While virtual screening of libraries of chemical compounds is not new, Coskunuzer said his group’s approach significantly outperforms other state-of-the-art methods on large data sets. More

Cornell University researchers have created an interface that allows users to handwrite and sketch within computer code — a challenge to conventional coding, which typically relies on typing.
The pen-based interface, called Notate, lets users of computational, digital notebooks open drawing canvases and handwrite diagrams within lines of traditional, digitized computer code.
Powered by a deep learning model, the interface bridges handwritten and textual programming contexts: notation in the handwritten diagram can reference textual code and vice versa. For instance, Notate recognizes handwritten programming symbols, like “n,” and then links them up to their typewritten equivalents.
“A system like this would be great for data science, specifically with sketching plots and charts that then inter-operate with textual code,” said Ian Arawjo, lead author of the paper and doctoral student in the field of information science. “Our work shows that the current infrastructure of programming is actually holding us back. People are ready for this type of feature, but developers of interfaces for typing code need to take note of this and support images and graphical interfaces inside code.”
Arawjo also said the work demonstrates a new path forward by introducing artificial intelligence-powered, pen-based coding at a time when drawing tablets are becoming more widely used.
“Tools like Notate are important because they open us up to new ways to think about what programming is, and how different tools and representational practices can change that perspective,” said Tapan Parikh, associate professor of information science and paper co-author.
Story Source:
Materials provided by Cornell University. Original written by Louis DiPietro. Note: Content may be edited for style and length. More

From pangolin scales that can stand up to hard hits to colorful but sturdy peacock feathers, nature can do a lot with a few simple molecules.
In a new review paper, a team of international researchers have laid out how engineers are taking inspiration from the biological world — and designing new kinds of materials that are potentially tougher, more versatile and more sustainable than what humans can make on their own.
“Even today, nature makes things way simpler and way smarter than what we can do synthetically in the lab,” said Dhriti Nepal, first author and a research materials engineer at the Air Force Research Laboratory in Ohio.
Nepal along with Vladimir Tsukruk from Georgia Institute of Technology and Hendrik Heinz of the University of Colorado Boulder served as co-corresponding authors for the new analysis. The team published its findings Nov. 28 in the journal Nature Materials.
The researchers, who come from three countries, delve into the promise and challenges behind “bioinspired nanocomposites.” These materials mix together different kinds of proteins and other molecules at incredibly small scales to achieve properties that may not be possible with traditional metals or plastics. Researchers often design them using advanced computer simulations or models. Examples include thin films that resist wear and tear by incorporating proteins from silkworm cocoons; new kinds of laminates made from polymers and clay materials; carbon fibers produced using bioinspired principles; and panes of glass that don’t easily crack because they include nacre — the iridescent lining inside many mollusk shells.
Such nature-inspired materials could, one day, lead to new and better solar panels, soft robots and even coatings for hypersonic jets, said Heinz, professor of chemical and biological engineering at CU Boulder. But first, researchers will need to learn how to build them from the bottom up, ensuring that every molecule is in the right place. More

Researchers affiliated with the Q-NEXT quantum research center show how to create quantum-entangled networks of atomic clocks and accelerometers — and they demonstrate the setup’s superior, high-precision performance.
What happened
For the first time, scientists have entangled atoms for use as networked quantum sensors, specifically, atomic clocks and accelerometers.
The research team’s experimental setup yielded ultraprecise measurements of time and acceleration. Compared to a similar setup that does not draw on quantum entanglement, their time measurements were 3.5 times more precise, and acceleration measurements exhibited 1.2 times greater precision.
The result, published in Nature, is supported by Q-NEXT, a U.S. Department of Energy (DOE) National Quantum Information Science Research Center led by DOE’s Argonne National Laboratory. The research was conducted by scientists at Stanford University, Cornell University and DOE’s Brookhaven National Laboratory.
“The impact of using entanglement in this configuration was that it produced better sensor network performance than would have been available if quantum entanglement were not used as a resource,” said Mark Kasevich, lead author of the paper, a member of Q-NEXT, the William R. Kenan, Jr. professor in the Stanford School of Humanities and Sciences and professor of physics and of applied physics. “For atomic clocks and accelerometers, ours is a pioneering demonstration.”
What is quantum entanglement? How does it apply to sensors? Entanglement, a special property of nature at the quantum level, is a correlation between two or more objects. When two atoms are entangled, one can measure the properties of both atoms by observing only one. This is true no matter how much distance — even if it’s light-years — separates the entangled atoms. A helpful everyday analogy: A red marble and a blue marble are placed in a box. If you draw a red marble from the box, you know, without having to look at the other one, that it’s blue. The color of the marbles is correlated, or entangled. In the quantum realm, entanglement is subtler. An atom can take on multiple states (colors) at once. If our marbles were like atoms, each marble would be both red and blue at the same time. Neither is fully red or blue while it sits the box. The quantum marble “decides” its color only at the moment of revelation. And once you draw one marble of “decided” color, you know the color of its entangled partner. To take a measurement of one member of an entangled pair is effectively to take a simultaneous reading of both. Taking this further: Two entangled clocks are practically equivalent to a single clock with two displays. Time measurements taken using entangled clocks can be more precise than measurements from two separate, synchronized clocks. More

Nanoengineers at the University of California San Diego’s Jacobs School of Engineering have developed an AI algorithm that predicts the structure and dynamic properties of any material — whether existing or new — almost instantaneously. Known as M3GNet, the algorithm was used to develop matterverse.ai, a database of more than 31 million yet-to-be-synthesized materials with properties predicted by machine learning algorithms. Matterverse.ai facilitates the discovery of new technological materials with exceptional properties.
The team behind M3GNet, led by UC San Diego nanoengineering professor Shyue Ping Ong, uses matterverse.ai and the new capabilities of M3GNet in their search for safer and more energy-dense electrodes and electrolytes for rechargeable lithium-ion batteries. The project is explored in the Nov. 28 issue of the journal Nature Computational Science.
The properties of a material are determined by the arrangement of its atoms. However, existing approaches to obtain that arrangement are either prohibitively expensive or ineffective for many elements.
“Similar to proteins, we need to know the structure of a material to predict its properties.” said Ong, the associate director of the Sustainable Power and Energy Center at the Jacobs School of Engineering. “What we need is an AlphaFold for materials.”
AlphaFold is an AI algorithm developed by Google DeepMind to predict protein structure. To build the equivalent for materials, Ong and his team combined graph neural networks with many-body interactions to build a deep learning architecture that works universally, with high accuracy, across all the elements of the periodic table.
“Mathematical graphs are really natural representations of a collection of atoms,” said Chi Chen, a former senior project scientist in Ong’s lab and first author of the work, who is now a senior quantum architect at Microsoft Quantum. “Using graphs, we can represent the full complexity of materials without being subject to the combinatorial explosion of terms in traditional formalisms.”
To train their model, the team used the huge database of materials energies, forces and stresses collected in the Materials Project over the past decade. The result is the M3GNet interatomic potential (IAP), which can predict the energies and forces in any collection of atoms. Matterverse.ai was generated through combinatorial elemental substitutions on more than 5,000 structural prototypes in the Inorganic Crystal Structure Database (ICSD). The M3GNet IAP was then used to obtain the equilibrium crystal structure — a process called “relaxation” — for property prediction.
Of the 31 million materials in matterverse.ai today, more than a million are predicted to be potentially stable. Ong and his team intend to greatly expand not just the number of materials, but also the number of ML-predicted properties, including high-value properties with small data sizes using a multi-fidelity approach they developed earlier.
Beyond structural relaxations, the M3GNet IAP also has broad applications in dynamic simulations of materials and property predictions as well.
“For instance, we are often interested in how fast lithium ions diffuse in a lithium-ion battery electrode or electrolyte. The faster the diffusion, the more quickly you can charge or discharge a battery,” Ong said. “We have shown that the M3GNet IAP can be used to predict the lithium conductivity of a material with good accuracy. We truly believe that the M3GNet architecture is a transformative tool that can greatly expand our ability to explore new material chemistries and structures.”
To promote the use of M3GNet, the team has released the framework as an open-source Python code on Github. Since posting the preprint on Arxiv in Feb 2022, the team has received interest from academic researchers and those in the industry. There are plans to integrate the M3GNet IAP as a tool in commercial materials simulation packages.
This work was authored by Chi Chen and Shyue Ping Ong at UC San Diego. The research was primarily funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, Materials Sciences and Engineering Division under the Materials Project program. Part of the work was funded by LG Energy Solution through the Frontier Research Laboratory Program. This work used the Extreme Science and Engineering Discovery Environment (XSEDE).
Story Source:
Materials provided by University of California – San Diego. Original written by Emerson Dameron. Note: Content may be edited for style and length. More

Researchers have developed a synthetic system that responds to environmental changes in the same way as living organisms, using a feedback loop to maintain its internal conditions. This not only keeps the material’s conditions stable but also makes it possible to build mechanisms that react dynamically to their environment, an important trait for interactive materials and soft robotics.
Living systems, from individual cells up to organisms, use feedback systems to maintain their conditions. For example, we sweat to cool down when we’re too warm, and a variety of systems work to keep our blood pressure and chemistry in the right range. These homeostatic systems make living organisms robust by enabling them to cope with changes in their environment. While feedback is important in some artificial systems, such as thermostats, they don’t have the dynamic adaptability or robustness of homeostatic living systems.
Now, researchers at Aalto University and Tampere University have developed a system of materials that maintains its state in a manner similar to living systems. The new system consists of two side-by-side gels with different properties. Interactions between the gels make the system respond homeostatically to environmental changes, keeping its temperature within a narrow range when stimulated by a laser.
‘The tissues of living organisms are typically soft, elastic and deformable,’ says Hang Zhang, an Academy of Finland postdoctoral researcher at Aalto who was one of the lead authors of the study. ‘The gels used in our system are similar. They are soft polymers swollen in water, and they can provide a fascinating variety of responses upon environmental stimuli.’
The laser shines through the first gel and then bounces off a mirror onto the second gel, where it heats suspended gold nanoparticles. The heat moves through the second gel to the first, raising its temperature. The first gel is only transparent when it is below a specific temperature; once it gets hotter, it becomes opaque. This change stops the laser from reaching the mirror and heating the second gel. The two gels then cool down until the first becomes transparent again, at which point the laser passes through and the heating process starts again.
In other words, the arrangement of the laser, gels and mirror creates a feedback loop that keeps the gels at a specific temperature. At higher temperatures, the laser is blocked and can’t heat the gold nanoparticles; at lower temperatures, the first gel becomes transparent, so the laser shines through and heats the gold particles.
‘Like a living system, our homeostatic system is dynamic. The temperature oscillates around the threshold, but the range of the oscillation is pretty small and is robust to outside disturbances. It’s a robust homeostatic system,’ says Hao Zeng, an Academy of Finland research fellow at Tampere University who was the other lead author of the study.
The researchers then built touch-responsive triggers on top of the feedback system. To accomplish this, they added mechanical components that respond to changes in temperature. Touching the gel system in the right way pushes it out of its steady state, and the resulting change in temperature causes the mechanical component to deform. Afterwards, everything returns to its original condition.
The team designed two systems that respond to different types of touch. In one case, a single touch triggers the response, just as a touch-me-not mimosa plant folds its leaves when stroked. The second setup only responds to repeated touches, in the same way as a Venus flytrap needs to be touched twice in 30 seconds to make it snap shut. ‘We can trigger a snapping behaviour with mechanical touches at suitable intervals, just like a Venus flytrap. Our artificial material system can discriminate between low-frequency and high-frequency touches,’ explains Professor Arri Priimägi of Tampere University.
The researchers also showed how the homeostatic system could control a dynamic colour display or even push cargo along its body. They emphasize that these demonstrations showcase only a handful of the possibilities opened up by the new material concept.
‘Life-inspired materials offer a new paradigm for dynamic and adaptive materials which will likely attract researchers for years to come,’ says Professor Olli Ikkala of Aalto University. ‘Carefully designed systems that mimic some of the basic behaviours of living systems will pave the way for truly smart materials and interactive soft robotics.’
Story Source:
Materials provided by Aalto University. Note: Content may be edited for style and length. More

Engineers at the Tokyo Institute of Technology (Tokyo Tech) have uncovered some intricate effects arising when chaotic systems, which typically generate broad spectra, are coupled by conveying only a narrow range of frequencies from one to another. The synchronization of chaotic oscillators, such as electronic circuits, continues to attract considerable fascination due to the richness of the complex behaviors that can emerge. Recently, hypothetical applications in distributed sensing have been envisaged, however, wireless couplings are only practical over narrow frequency intervals. The proposed research shows that, even under such constraints, chaos synchronization can occur and give rise to phenomena that could one day be leveraged to realize useful operations over ensembles of distant nodes.
The abstract notion that the whole can be found in each part of something has for long fascinated thinkers engaged in all walks of philosophy and experimental science: from Immanuel Kant on the essence of time to David Bohm on the notion of order, and from the self-similarity of fractal structures to the defining properties of holograms. It has, however, remained understandably extraneous to electronic engineering, which strives to develop ever more specialized and efficient circuits exchanging signals that possess highly controlled characteristics. By contrast, across the most diverse complex systems in nature, such as the brain, the generation of activity having features that present themselves similarly over different temporal scales, or frequencies, is nearly a ubiquitous observation.
In a quest to explore new and unorthodox approaches to designing systems capable of solving difficult computation and control problems, physicists and engineers have, for decades, been investigating networks made up of chaotic oscillators. These are systems that can be easily realized using analog electronic, optical, and mechanical components. Their striking property is that, despite being quite simple in their structure, they can generate behaviors that are, at the same time, incredibly intricate and far from random. “Chaos entails an extreme sensitivity to initial conditions, meaning that the activity at each point in time is effectively unpredictable. However, a crucial aspect is that the geometrical arrangements of the trajectories generated by chaotic signals have well-defined properties which, alongside the distribution of frequencies, are rather stable and repeatable. Since these features can change in many ways depending on the voltage input or parameter settings like a resistor value, these circuits are interesting as a basis for realizing new forms of distributed computation, for example, based on sensor readings,” explains Dr. Ludovico Minati, lead author of the study. “In our recent work, we showed that they could be effectively used to realize the kind of physical reservoirs that can simplify neural network training,” adds Mr. Jim Bartels, doctoral student at the Nano Sensing Unit, where the study was conducted [1].
When two or more chaotic oscillators are coupled together, the most interesting behaviors emerge as they attract and repulse their activities while trying to find an equilibrium, in ways that usual periodic oscillators simply cannot access. “Two years ago, work done in our laboratory demonstrated that these behaviors could, at least in principle, be used as a means to gather readings from distant sensors and directly provide statistics such as the average value,” adds Dr. Ludovico Minati [2]. However, the complex nature of chaotic signals implies that they generally feature broad frequency spectra, which are very different from those, narrow and neatly delineated, that are typically used in modern wireless communication. “As a consequence, it becomes very difficult, if not impossible, to realize couplings over the air. That’s not only because antennas are often highly tuned for specific frequencies, but also and especially because radio regulations do not allow broadcasting except within tightly-defined regions,” explains Mr. Boyan Li, master student and second author of the study.
To date, there is a substantial body of literature covering the many effects that can arise in ensembles of chaotic oscillators. For example, small groups of nodes that preferentially synchronize with each other can appear, a little like groups of people coalescing together at a party, together with unexpected remote inter-dependencies that remind us of the binding problem in the brain. However, surprisingly, almost no studies have considered the possibility (or otherwise) of coupling chaotic oscillators via a mechanism, basically a filter, that transfers only a narrow range of frequencies. For this reason, the researchers at Tokyo Tech decided to explore the behavior of a pair of chaotic oscillators. They coupled them using a filter that they could easily tune to let through only a narrow range of frequencies, while for the time being retaining a wired connection between them.
“We decided to use a type of chaotic oscillator that is extraordinarily simple, involving only one transistor and a handful of passive components, and known as the Minati-Frasca oscillator. This family of oscillators was introduced about five years ago by researchers from Italy and Poland, and has many remarkable properties, as outlined in a recent book. Recently, we become interested in understanding them and their several potential applications,” explains Dr. Hiroyuki Ito, head of the Nano Sensing Unit where the study was conducted.
Based on simulations and measurements, the research team was able to demonstrate that it is in fact possible to synchronize these oscillators even without transferring the entire broad spectrum, but just a relatively narrow “slice” of it. They like to compare this to a situation where the whole is found, at least partially, in a part. When operating in the lower gigahertz region, close to where first-generation wireless devices work, the oscillators could synchronize when conveying only a few point percent of the bandwidth. As expected, the synchronization was not complete, meaning that the oscillators did not completely follow each other’s activity. “This sort of incomplete, or weak, interdependence is precisely the region where the most interesting effects can appear at the level of a network of nodes. It is quite similar between oscillators and neurons, as one of our previous works showed. These are the mechanisms that represent the next frontier for implementing distributed computation based on emergent behaviors, as many research groups worldwide are pursuing,” adds Dr. Mattia Frasca from the University of Catania in Italy, who initially co-discovered these circuits with Dr. Minati, later analyzing together their behaviors and relationship to other systems in nature, and provided several theoretical foundations which were used for the study by the Tokyo Tech researchers.
The researchers observed that while a narrow slice of the spectrum was enough to obtain some detectable synchronization, the center location and width of the filter had important effects. Using a multitude of analysis techniques, they could see that over some regions, the activity of the slave oscillator tracked the filter setting in an evident way, whereas in others, different and rather more complex effects appeared. “This is a good example of the richness of behaviors available to these circuits, which remain not widely known in the electronic engineering community. It is quite different compared to the simpler responses of periodic systems, which are either locked or not to each other. It is a long way before we are really able to realize effective applications using these phenomena, so it must be said that this is fundamental research at the moment. However, it is very fascinating to think that in the future we may realize some aspects of sensing also using these unusual approaches,” adds Ms. Zixuan Li, doctoral student, and co-author of the study.
After this interview, the team explained that this type of research will firstly need to be extended by understanding more deeply the phenomena and how they can be used to generate interesting collective activity. Then, the two main engineering challenges will be to demonstrate couplings over an actual wireless link, while meeting all radio requirements, and to substantially minimizing the power consumption, also using some results from their previous research. “If successful solutions are found to these challenges, then one of our main goals is to demonstrate usable distributed sensing in applications that are important to society, such as monitoring the condition of land in precision agriculture,” concludes Dr. Hiroyuki Ito. The methodology and results are reported in a recent article published in the journal Chaos, Solitons and Fractals [3], and all of the experimental recordings have been made freely available for others to use in future work. More