The impending danger of bacterial resistance to antibiotics is well-documented within the scientific community. Bacteria are the most efficient evolvers, and their ability to develop tolerance to drugs, in addition to antibiotic overuse and misuse, means that researchers have had to get particularly resourceful to ensure the future of modern medicine.
WIRED’s Max G. Levy recently spoke with de la Fuente and postdoctoral researcher and study collaborator Marcelo Torres about the urgency of the team’s work, and why developing these solutions is critical to the survival of civilization as we know it. The team’s algorithm, based on pattern recognition software used to analyze images, makes an otherwise insurmountable feat tangible.
De la Fuente’s lab specializes in using AI to discover and design new drugs. Rather than making some all-new peptide molecules that fit the bill, they hypothesized that an algorithm could use machine learning to winnow down the huge repository of natural peptide sequences in the human proteome into a select few candidates.
“We know those patterns—the multiple patterns—that we’re looking for,” says de la Fuente. “So that allows us to use the algorithm as a search function.”
While biologists and chemists race to develop new antibiotics to combat constantly mutating bacteria, predicted to lead to 10 million deaths by 2050, engineers are approaching the problem through a different lens: finding naturally occurring antibiotics in the human genome.
The billions of base pairs in the genome are essentially one long string of code that contains the instructions for making all of the molecules the body needs. The most basic of these molecules are amino acids, the building blocks for peptides, which in turn combine to form proteins. However, there is still much to learn about how — and where — a particular set of instructions are encoded.
Now, bringing a computer science approach to a life science problem, an interdisciplinary team of Penn researchers have used a carefully designed algorithm to discover a new suite of antimicrobial peptides, hiding deep within this code.
The study, published in Nature Biomedical Engineering, was led by César de la Fuente, Presidential Assistant Professor in Bioengineering, Microbiology, Psychiatry, and Chemical and Biomolecular Engineering, spanning both Penn Engineering and Penn Medicine, and his postdocs Marcelo Torres and Marcelo Melo. Collaborators Orlando Crescenzi and Eugenio Notomista of the University of Naples Federico II also contributed to this work.
“The human body is a treasure trove of information, a biological dataset. By using the right tools, we can mine for answers to some of the most challenging questions,” says de la Fuente. “We use the word ‘encrypted’ to describe the antimicrobial peptides we found because they are hidden within larger proteins that seem to have no connection to the immune system, the area where we expect to find this function.”
While artificial intelligence is becoming a bigger part of nearly every industry and increasingly present in everyday life, even the most impressive AI is no match for a toddler, chimpanzee, or even a honeybee when it comes to learning, creativity, abstract thinking or connecting cause and effect in ways they haven’t been explicitly programmed to recognize.
This discrepancy gets at one of the field’s fundamental questions: what does it mean to say an artificial system is “intelligent” in the first place?
Seventy years ago, Alan Turing famously proposed such a benchmark; a machine could be considered to have artificial intelligence if it could successfully fool a person into thinking it was a human as well. Now, many artificial systems could pass a “Turing Test” in certain limited domains, but none come close to imitating the holistic sense of intelligence we recognize in animals and people.
Understanding how AI might someday be more like this kind of biological intelligence — and developing new versions of the Turing Test with those principles in mind — is the goal of a new collaboration between researchers at the University of Pennsylvania, Carnegie Mellon University and Johns Hopkins University.
The project, called “From Biological Intelligence to Human Intelligence to Artificial General Intelligence,” is led by Konrad Kording, a Penn Integrates Knowledge Professor with appointments in the Departments of Bioengineering and Computer and Information Science in Penn Engineering and the Department of Neuroscience at Penn’s Perelman School of Medicine. Kording will collaborate with Timothy Verstynen of Carnegie Mellon University, as well Joshua T. Vogelstein and Leyla Isik, both of Johns Hopkins University, on the project.
Diabetes is one of the more common diseases among Americans today, with the American Diabetes Association estimating that approximately 9.5 percent of the population battles the condition today. Though symptoms and causes may vary across types and patients, diabetes generally results from the body’s inability to produce enough insulin to keep blood sugar levels in check. A new experimental treatment from the lab of Sha Jin, Ph.D., a biomedical engineering professor at Binghamton University, aims to use about $1.2 million in recent federal grants to develop a method for pancreatic islet cell transplantation, as those are the cells responsible for producing insulin.
But the catch to this new approach is that relying on healthy donors of these islet cells won’t easily meet the vast need for them in diabetic patients. Sha Jin wants to use her grants to consider the molecular mechanisms that can lead pluripotent stem cells to become islet-like organoids. Because pluripotent stem cells have the capability to evolve into nearly any kind of cell in the human body, the key to Jin’s research is learning how to control their mechanisms and signaling pathways so that they only become islet cells. Jin also wants to improve the eventual culture of these islet cells into three-dimensional scaffolds by finding ways of circulating appropriate levels of oxygen to all parts of the scaffold, particularly those at the center, which are notoriously difficult to accommodate. If successful in her tissue engineering efforts, Jin will not only be able to help diabetic patients, but also open the door to new methods of evolving pluripotent stem cells into mini-organ models for clinical testing of other diseases as well.
A Treatment to Help Others See Better
Permanently crossed eyes, a medical condition called strabismus, affects almost 18 million people in the United States, and is particularly common among children. For a person with strabismus, the eyes don’t line up to look at the same place at the same time, which can cause blurriness, double vision, and eye strain, among other symptoms. Associate professor of bioengineering at George Mason University, Qi Wei, Ph.D., hopes to use almost $2 million in recent funding from the National Institute of Health to treat and diagnose strabismus with a data-driven computer model of the condition. Currently, the most common method of treating strabismus is through surgery on one of the extraocular muscles that contribute to it, but Wei wants her model to eventually offer a noninvasive approach. Using data from patient MRIs, current surgical procedures, and the outcomes of those procedures, Wei hopes to advance and innovate knowledge on treating strabismus.
A Newly Analyzed Brain Mechanism Could be the Key to Stopping Seizures
Among neurological disorders, epilepsy is one of the most common. An umbrella term for a lot of different seizure-inducing conditions, many versions of epilepsy can be treated pharmaceutically. Some, however, are resistant to the drugs used for treatment, and require surgical intervention. Bin He, Ph. D., the Head of the Department of Biomedical Engineering at Carnegie Mellon University, recently published a paper in collaboration with researchers at Mayo Clinic that describes the way that seizures originating at a single point in the brain can be regulated by what he calls “push-pull” dynamics within the brain. This means that the propagation of a seizure through the brain relies on the impact of surrounding tissue. The “pull” he refers to is of the surrounding tissue towards the seizure onset zone, while the “push” is what propagates from the seizure onset zone. Thus, the strength of the “pull” largely dictates whether or not a seizure will spread. He and his lab looked at different speeds of brain rhythms to perform analysis of functional networks for each rhythm band. They found that this “push-pull” mechanism dictated the propagation of seizures in the brain, and suggest future pathways of treatment options for epilepsy focused on this mechanism.
Hyperspectral Imaging Might Provide New Ways of Finding Cancer
A new method of imaging called hyperspectral imaging could help improve the prediction of cancerous cells in tissue specimens. A recent study from a University of Texas Dallas team of researchers led by professor of bioengineering Baowei Fei, Ph.D., found that a combination of hyperspectral imaging and artificial intelligence led to an 80% to 90% level of accuracy in identifying the presence of cancer cells in a sample of 293 tissue specimens from 102 patients. With a $1.6 million grant from the Cancer Prevention and Research Institute of Texas, Fei wants to develop a smart surgical microscope that will help surgeons better detect cancer during surgery.
Fei’s use of hyperspectral imaging allows him to see the unique cellular reflections and absorptions of light across the electromagnetic spectrum, giving each cell its own specific marker and mode of identification. When paired with artificial intelligence algorithms, the microscope Fei has in mind can be trained to specifically recognize cancerous cells based on their hyperspectral imaging patterns. If successful, Fei’s innovations will speed the process of diagnosis, and potentially improve cancer treatments.
People and Places
A group of Penn engineering seniors won the Pioneer Award at the Rothberg Catalyzer Makerthon led be Penn Health-Tech that took place from October 19-20, 2019. SchistoSpot is a senior design project created by students Vishal Tien (BE ‘20), Justin Swirbul (CIS ‘20), Alec Bayliff (BE ‘20), and Bram Bruno (CIS ‘20) in which the group will design a low-cost microscopy dianostic tool that uses computer vision capabilities to automate the diagnosis of schistosomiasis, which is a common parasitic disease. Read about all the winners here.
Virginia Tech University will launch a new Cancer Research Initiative with the hope of creating an intellectual community across engineers, veterinarians, biomedical researchers, and other relevant scientists. The initiative will focus not only on building better connections throughout departments at the university, but also in working with local hospitals like the Carilion Clinic and the Children’s National Hospital in Washington, D.C. Through these new connections, people from all different areas of science and engineering and come together to share ideas.
Associate Professor of Penn Bioengineering Dani Bassett, Ph.D., recently sat down with the Penn Integrates Knowledge University Professor Duncan Watts, Ph.D., for an interview published in Penn Engineering. Bassett discusses the origins of network science, her research in small-world brain networks, academic teamwork, and the pedagogy of science and engineering. You can read the full interview here.
DNA Microscopy Gives a Better Look at Cell and Tissue Organization
A new technique that researchers from the Broad Institute of MIT and Harvard University are calling DNA microscopy could help map cells for better understanding of genetic and molecular complexities. Joshua Weinstein, Ph.D., a postdoctoral associate at the Broad Institute, who is also an alumnus of Penn’s Physics and Biophysics department and former student in Penn Bioengineering Professor Ravi Radhakrishnan’s lab, is the first author of this paper on optics-free imaging published in Cell.
The primary goal of the study was to find a way of improving analysis of the spatial organization of cells and tissues in terms of their molecules like DNA and RNA. The DNA microscopy method that Weinstein and his team designed involves first tagging DNA, and allowing the DNA to replicate with those tags, which eventually creates a cloud of sorts that diffuses throughout the cell. The DNA tags subsequent interactions with molecules throughout the cell allowed Weinstein and his team to calculate the locations of those molecules within the cell using basic lab equipment. While the researchers on this project focused their application of DNA microscopy on tracking human cancer cells through RNA tags, this new method opens the door to future study of any condition in which the organization of cells is important.
Penn Engineers Demonstrate Superstrong, Reversible Adhesive that Works like Snail Slime
If you’ve ever pressed a picture-hanging strip onto the wall only to realize it’s slightly off-center, you know the disappointment behind adhesion as we typically experience it: it may be strong, but it’s mostly irreversible. While you can un-stick the used strip from the wall, you can’t turn its stickiness back on to adjust its placement; you have to start over with a new strip or tolerate your mistake. Beyond its relevance to interior decorating, durable, reversible adhesion could allow for reusable envelopes, gravity-defying boots, and more heavy-duty industrial applications like car assembly.
Such adhesion has eluded scientists for years but is naturally found in snail slime. A snail’s epiphragm — a slimy layer of moisture that can harden to protect its body from dryness — allows the snail to cement itself in place for long periods of time, making it the ultimate model in adhesion that can be switched on and off as needed. In a new study, Penn Engineers demonstrate a strong, reversible adhesive that uses the same mechanisms that snails do.
Low-Dose Radiation CT Scans Could Be Improved by Machine Learning
Machine learning is a type of artificial intelligence growing more and more popular for applications in bioengineering and therapeutics. Based on learning from patterns in a way similar to the way we do as humans, machine learning is the study of statistical models that can perform specific tasks without explicit instructions. Now, researchers at Rensselaer Polytechnic Institute (RPI) want to use these kinds of models in computerized tomography (CT) scanning by lowering radiation dosage and improving imaging techniques.
A recent paper published in Nature Machine Intelligence details the use of modularized neural networks in low-dose CT scans by RPI bioengineering faculty member Ge Wang, Ph.D., and his lab. Since decreasing the amount of radiation used in a scan will also decrease the quality of the final image, Wang and his team focused on a more optimized approach of image reconstruction with machine learning, so that as little data as possible would be altered or lost in the reconstruction. When tested on CT scans from Massachusetts General Hospital and compared to current image reconstruction methods for the scans, Wang and his team’s method performed just as well if not better than scans performed without the use of machine learning, giving promise to future improvements in low-dose CT scans.
A Mind-Controlled Robotic Arm That Requires No Implants
A new mind-controlled robotic arm designed by researchers at Carnegie Mellon University is the first successful noninvasive brain-computer interface (BCI) of its kind. While BCIs have been around for a while now, this new design from the lab of Bin He, Ph.D., a Trustee Professor and the Department Head of Biomedical Engineering at CMU, hopes to eliminate the brain implant that most interfaces currently use. The key to doing this isn’t in trying to replace the implants with noninvasive sensors, but in improving noisy EEG signals through machine learning, neural decoding, and neural imaging. Paired with increased user engagement and training for the new device, He and his team demonstrated that their design enhanced continuous tracking of a target on a computer screen by 500% when compared to typical noninvasive BCIs. He and his team hope that their innovation will help make BCIs more accessible to the patients that need them by reducing the cost and risk of a surgical implant while also improving interface performance.
KIChE is an organization that aims “to promote constructive and mutually beneficial interactions among Korean Chemical Engineers in the U.S. and facilitate international collaboration between engineers in U.S. and Korea.”
We would also like to congratulate Natalia Trayanova, Ph.D., of the Department of Biomedical Engineering at Johns Hopkins University on being inducted into the Women in Tech International (WITI) Hall of Fame. Beginning in 1996, the Hall of Fame recognizes significant contributions to science and technology from women. Trayanova’s research specializes in computational cardiology with a focus on virtual heart models for the study of individualized heart irregularities in patients. Her research helps to improve treatment plans for patients with cardiac problems by creating virtual simulations that help reduce uncertainty in either diagnosis or courses of therapy.
T lymphocytes in the immune system play a vital role in the body to recognize invasion by an outside element. When foreign bacteria enter the body, receptors on the T cell surface detect antigens associated with the bacteria and send a signal deploying phagocytes to attack and defeat the invading bacteria. While evolution and vaccination make the immune system very efficient, the inability of T cell receptors (TCRs) to detect cancer makes normal T cells relatively ineffective in resisting cancer. One of the ways to overcome this limitation of the immune system is to better understand how the TCRs respond to antigens. Analyses of the proteins involved in TSR responses are useful but limited by several factors, including the dizzying amount of data involved. Data analysis techniques have been helpful but have offered little information about the general reactions of TSRs, rather than how they react to specific antigens.
A possible solution to this obstacle is ImmunoMap, developed by scientists collaborating between Johns Hopkins University and Memorial Sloan Kettering Cancer Center. In a study recently published in Cancer Immunology Research, the authors, led by Jonathan P. Schneck, M.D., Ph.D., a professor of pathology at Johns Hopkins associated with the university’s Institute for Cell Engineering and Institute for Nanobiotechnology, describe their creation and deployment of ImmunoMap, a group of artificial intelligence algorithms that use machine learning to process large amounts of sequencing data and compare data from different antigens with each other.
The authors trained ImmunoMap initially using data from a mouse model of melanoma, in which the algorithm demonstrated significantly better performance than traditional methods. Subsequently, ImmunoMap was applied to patient response data from a melanoma clinical trial of the chemotherapy agent nivolumab. The algorithm discovered a new group of patients that would respond positively to nivolumab treatment — a finding missed by popular past methods. More testing of ImmunoMap is necessary, but the technology could make significant contributions to the monitoring of cancer patients receiving chemotherapy. In addition, it could to help to better predict response in patients before they begin specific chemotherapy regimes.
Wearables Improving Health
Among the most troubling health disparities related to global wealth inequality is the higher rate of mortality among children suffering from cancer. Fever is a common symptom of children undergoing cancer treatment, and this symptom may indicate more serious health issues that require the attention of a doctor. However, continuously monitoring skin temperature in children from low resource settings is difficult. Seeking to help remedy this problem, undergraduate engineering students at Harvard collaborated with the Dana-Farber/Boston Children’s Cancer & Blood Disorders Center’s Global Health Initiative to develop tools for earlier fever detection and treatment.
In a course taught by David Mooney, Ph.D., Robert P. Pinkas Family Professor of Bioengineering at Harvard, students developed a wearable device that sounds an alarm when the wearer needs medical help. The app can send patients’ recorded messages to their doctors, who can then review the temperature data and messages from the children before responding. Fashioned like a wristwatch, the extra-durable and waterproof device will next move into pilot testing among a larger patient population.
Meanwhile, at Northwestern, John A. Rogers, Ph.D., the Louis Simpson and Kimberly Querrey Professor of Materials Science and Engineering, Biomedical Engineering, and Neurological Surgery in Northwestern’s McCormick School of Engineering, has partnered with cosmetics giant L’Oréal to create the world’s smallest wearable. The device, which is smaller than an adult fingernail, measures UV sun exposure for the wearer and can tell when they should go back inside instead of risking overexposure. Unsurprisingly, it’s solar powered, and it was demonstrated a couple of weeks ago at a consumer electronics show in Las Vegas.
Growing Hydrogels Like Human Tissue
Scientists at Carnegie Mellon University and Nanyang Technological University in Singapore have collaborated in a process to create polyacrylamide gels that grow in a manner resembling natural tissue. K. Jimmy Hsia, Ph.D., Professor of Biomedical Engineering at Carnegie Mellon, is co-lead author of a new study in PNAS describing this new growth mode.
In the study, Dr. Hsia and his coauthors report that, in the same way that growth factors secreted by a living organism affect the generation of new tissue, oxygen can be modulated to control how hydrogels grow. Moreover, while growth is under way, the process could be continued to efficiently manage the mass transfer of nutrients from cell to cell. Finally, the authors detail the mechanical processes that help to shape the final product. With this new process, the ability to design and create materials for applications such as robotics and tissue engineering comes a step closer to resembling living tissue as closely as possible.
People and Places
Engineers at Virginia Tech have been awarded a $1.1 million grant from the Virginia Research Investment Committee to develop a device that uses low-energy electric fields for the treatment of brain tumors. Rafael Davalos, Ph.D., L. Preston Wade Professor of Biomedical Engineering and Mechanics, is the chief investigator on the grant.
The Department of Biomedical Engineering has announced the appointment of Kam W. Leong, Ph.D., as the Samuel Y. Sheng Professor of Biomedical Engineering. Dr. Leong earned his Ph.D. in chemical engineering from the University of Pennsylvania and taught at Duke and Johns Hopkins before arriving at Columbia in 2006. He was previously the James B. Duke Professor of Biomedical Engineering at Duke. Congratulations to Dr. Leong!