Research Blog

In April of 2019, the first government declared climate change to be a national emergency. Since then, over 1,900 local governments and more than 23 national governments have expressed the same sentiment.

A 2021 report released by the IPCC labeled climate change a ‘code red’ for humanity, and every day more than 2 million people search the term ‘climate crisis’ on Google. So it’s apparent, the climate crisis is imminent. What’s the solution? Experts at Duke’s annual Research Week posed their research-based solutions during a virtual panel hosted on February 1st. (View the Session)

Mark Borsuk and Jonathan Wiener’s analysis of risk factors concerning their solution, SRM.

The panel, mediated by Biology professor Mohamed Noor, began with a solution posed by professors Mark Borsuk and Jonathan Wiener. Known as solar radiation modification, SRM is “an attempt to moderate global warming by intentionally increasing the amount of incoming sunlight that is reflected by the atmosphere back to space,” according to Borsuk. Its primary technique is stratospheric aerosol injection. Wiener explained that their research is “trying to understand the risk… And we’re working to study these multiple impacts because all too often, as we’re all familiar with human decision making at the individual level or the governmental public policy level tends to focus on one thing at a time.” However, even with possible governance challenges at play, their research poses an extremely cheap yet effective solution for avoiding some of the worst impacts of climate change.

Dalia Patino Echeverri’s presentation on GRACE, an energy solution.

Next up on the panel was Dalia Patino Echeverri, an associate professor at Duke’s Nicholas School of the Environment. She began by ruminating on the challenges faced in Texas after the snowstorm last year, and how climate change intensified those challenges. Her research focuses on how to address the electricity issues that climate change is producing in our nation, through a system called ‘GRACE’. ‘GRACE’ is a power grid that is risk-aware for clean, smart energy usage.

“It considers the forecast of electricity, the amount of load on the forecast of electricity generation from wind and sun of resources, and looks at the availability of conventional resources to schedule this commercial resources.” said Echeverri. Its operating system is extremely intelligent minimizing expected value and total cost of energy during times of climate crisis.

Brian Silliman’s presentation on Duke Restore.

Finally, a solution was presented from Brian Silliman, the Rachel Carson Distinguished Professor for marine biology. He introduced a more grassroots approach to climate restoration, called Duke Restore.

“A lot of our research and those of others have shown that the presence of restored marine environments greatly protects human societies on the coastline from increasing threat storm surge, and flooding generated in large part by climate change impacts, etc.” Silliman began.

Duke Restore aims to go out into ecosystems and restore the shorelines that have been lost, indirectly aiding in climate crisis alleviation. Silliman is currently collaborating with governments and other conservation organizations to help change the way they plan to restore these ecosystems from the bottom up. ““We’re doing this here in North Carolina with the US Marine Corps, changing the planting designs to switch the restoration trajectory from failure to success.”

Kay Jowers explaining her ideas for a more equitable approach to policy solutions.

Kay Jowers, a Senior Policy Associate at the Duke Nicholas Institute for Environmental Policy Solutions, closed out the panel event with some final thoughts.

“My charge is to give you some food for thought about creating a more supportive environment for environmental and climate justice at Duke,” she began. She explained the need for action as compared to documentation and explained that equitable approaches are needed to avoid a climate disaster.

“In the world of Environmental Justice Studies, the communities, and the scholars have been calling for less problematization and documenting of problems, and more orientation towards solutions.” Her sentiments resonated deeply with the theme of the panel, as solution-based research is of paramount importance in the 21st century.

The Duke Research Week panel on climate change solutions posed tangible explications for the ever imminent climate crisis happening around the world. Though climate change is apparent now more than ever, researchers like these hold the solutions for the future.

Post by Skylar Hughes, Class of 2025

Ethereum: What are Transaction Fees and How are They Determined?

By Anna Gotskind

On February 2, 2022

In Computers/Technology, Data, Engineering, Students

By now most people have heard of Bitcoin, the first form of decentralized cryptocurrency which was created in 2009 and popularized in 2011. However, these novel tokens did not just appear out of thin air, they had to be mined. But what does this mean?

Essentially, there is a finite amount of Bitcoin, 21 million to be exact. Bitcoin miners run complex computer rigs to solve intricate and complicated puzzles in order to confirm groups of bitcoin transactions called blocks. Once a block is mined, the miner is rewarded with bitcoin.

On 3 January 2009, the bitcoin network came into existence after the founder, Satoshi Nakamoto, mined the genesis block of bitcoin (block number 0), and received a reward of 50 bitcoins. The rewards for Bitcoin mining are reduced by half roughly every four years due to its scarcity. Currently, miners are rewarded 6.25 Bitcoins for every block. Additionally, when a transaction is approved via mining, it is added to a block which is then added to the Bitcoin blockchain. A blockchain is an immutable, decentralized, and transparent computer network that acts as a publicly available ledger. For more information please reference my previous article here.

Not all tokens are mined, however, the most popular or widely used ones, Bitcoin and Ethereum are. Today, we will be focusing on the Ethereum Blockchain using ETH tokens.

Similar to Bitcoin, ETH is also mined by solving complex puzzles in order to confirm and verify blockchain transactions. However, ETH miners are paid in ETH, not bitcoin. In addition to receiving the ETH from mining, miners are also paid through transaction fees called gas.

Transaction fees are determined by a Transaction fee mechanism (TFM), a key component of blockchain protocol. However, there has yet to be an empirical study on the real-world impact of TFMs. Recently, a study out of Duke and Peking University evaluated the effect of EIP-1559, the first TFM to abandon the traditional first-price auction paradigm.

Every transaction or smart contract executed on the Ethereum blockchain requires gas. If you are unfamiliar with smart contracts please reference my previous article here.

“Gas is a unit of measurement for the amount of computational effort required to execute a specific on-network operation”
William Zhao ’23, Student researcher

However, the price of gas is constantly changing in response to how many others are trying to make transactions on the blockchain. Gas prices are typically denoted in GWEI or a billionth of an ETH ( 0.000000001 ETH). For context as of February 1st, 2022 at 1:17 ET, ETH is worth $2778.50 USD per token.

When an ETH transaction is placed it is not immediately completed and resides in a memory pool or “Mempool.” These are smaller databases of unconfirmed or pending transactions. Prior to the EIP-1559 update, the Ethereum TFM centered around the first-price auction paradigm.

Conceptually, the first-price auction paradigm is fairly simple. Essentially every time a transaction is made there is an accompanying gas bid. Crypto wallets like Metamask or Coinbase Wallet provide suggested gas bids for users but still allow them to alter the bid. This is because transaction verification priority is determined by the miner and thus given to whoever bids the most. Once a transaction is verified it is added to the miner’s block and then to the blockchain. As a result, some users would offer unnecessarily high gas fees in order for their transaction to skip the line and be quickly processed thus creating major delays for others.

There were several problems under this previous TFM including long wait times for verification, extremely high gas and unpredictable prices, as well as inefficiencies around block size and consensus security. Recent research examined the causal effect of EIP-1559 on blockchain transaction fee dynamics, transaction waiting time, and security. They found that while the transaction mechanism became even more complex it did also become more efficient.

EIP-1559 improves user experience by reducing users’ waiting times, improving fee estimation, and mitigating intra-block difference of gas price paid (which is more important for miners). However, EIP-1559 did not have a large impact on gas fee reduction or consensus security. In addition, they found that when ETH’s price is more volatile, the waiting time is significantly higher.

Figure 8: Distributions of median waiting time. Users experience a much lower transaction waiting time following EIP-1559.

Ultimately, while user experience improved, scalability issues held the TFM from having a larger effect on important components like gas prices.

“If you can only hold a certain amount of transactions that’s a hard cap on development, however, high gas prices are a scalability issue not a mechanism design issue.
William Zhao ’23, student researcher

This research paper was recognized by Vitalik Buterin, one of the co-founders of Ethereum.

Remembrance of Wordles Past

By Sophie Cox

On February 1, 2022

In Computers/Technology, Data, Students

Devang Thakkar, a fourth-year PhD candidate at Duke University, recently created an archive for Wordle that gives users unlimited access to past Wordle games. Gray tiles indicate letters not found anywhere in the correct word, yellow indicates letters that are in the word but not in the right place, and green indicates correctly placed letters.

Writing this story was dangerous. Before, I was only vaguely aware of the existence of Wordle, a wildly popular online word game created by Josh Wardle and recently bought by the New York Times. Now I can’t stop playing it. The objective of the game sounds deceptively simple: try to guess the right five-letter word in six attempts or fewer.

Thanks to Devang Thakkar, a fourth-year PhD student in Computational Biology and Bioinformatics at Duke, the 200+ Wordle games released before I discovered its charms are readily accessible online. So now I’m making up for lost time.

Thakkar recently spent a weekend building an archive of every Wordle game in existence. You can play them in any order. You can start at the beginning. You can start with today’s Wordle and work backward. You can sit down and play eight in a row. Just hypothetically, of course.

Devang Thakkar became hooked on Wordle when his roommate introduced it to him, but he wanted a way to access old Wordles as well. First, he experimented with manually changing the date on his browser to trick the computer into showing him old Wordles. However, his browser gave him an error message if he tried to go back more than fourteen days. To get around that, Mr. Thakkar wrote a Python script using a Python library called Selenium, which allowed him “to basically go back as much as you want.”

Thakkar combined his own data with an open-source Wordle project called WordMaster created by Katherine Peterson. With an open-source project, Thakkar says, “You put your work out there, and then someone else adds to it.”

Devang Thakkar at the 2020 Data Through Design exhibition in New York.
Photograph courtesy of Devang Thakkar.

Whereas WordMaster randomly generates new five-letter words, Thakkar’s archive provides access to “official” Wordle games from the past. While there were many random Wordle generators already in existence, it was the usage of the official Wordle list and the ability to go back to a particular Wordle that set this archive apart. Thakkar also added features like the ability to share your answers with others and an option that lets users access Wordle games in a random order.

Thakkar tells me the project was “just for fun.” “I was bored… so I was like, ‘let’s make something!’” he says. Nevertheless, “That is essentially what I do for my work as well; I write code.” In the Dave Lab, Devang Thakkar uses sequencing data to study the origins of different types of lymphomas.

In his free time, Devang Thakkar enjoys woodworking and metalworking. Pictured here are two of his projects, a wooden bowl and his own dining room table.
Photographs courtesy of Devang Thakkar.

When he’s not working or making Wordle archives, Devang Thakkar can often be found in Duke’s Innovation Co-Lab, where he enjoys woodworking and metalworking. His projects range from creations intended as gifts, like a laptop stand and beer caddy, to his own dining room table. Thakkar says the hobby, being very different from his normal work, helps him maintain work-life balance.

The Wordle project, on the other hand, required coding skills Thakkar uses daily. “This is just like work for me, but for fun.” He enjoys graphic design and board games and has “a special affection for board games with words.”

As for the Wordle archive, Mr. Thakkar says he never expected it to become so popular. He thought it would mostly be used by his friends, but the archive quickly accumulated millions of weekly users. “People keep sending me screenshots of their friends sending them this website,” he says.

Meanwhile, I’ve started noticing Wordle references everywhere. Just after I spoke to Thakkar about his project, I happened to stumble across a link to BRDL, a delightful Wordle spinoff that uses four-letter birding codes instead of words. By blind luck, I guessed the right code on my second try: AMGO, American goldfinch. A few days later, I overheard two students talking about the daily Wordle. Clearly, I’m not the only one who’s become hooked on the game. Fortunately for everyone who is, Devang Thakkar’s Wordle archive, which he called “Remembrance of Wordles Past,” offers unlimited access.

Introducing Muser – A Better Way to Find Student Research

By Skylar Hughes

On January 26, 2022

In Field Research, Genetics/Genomics, History, Science Communication & Education, Students

An effortlessly simple research platform where Duke students and Duke research projects can connect? Yes, please!

If you are anything like me, Duke University’s incredible research opportunities were extremely enticing when considering this school. One of the top 10 research institutions in the United States, Duke University’s research community spends over 1 billion dollars annually to fund its projects, which includes notable research facilities like the Duke Center for Human Genetics, the Duke Cancer Institute, the Duke Center for AIDS Research, and the Duke Human Vaccine Institute.

However, the amount of opportunity in this area can be overwhelming to approach, and as a student you often have no clue where to start.

*Summer undergraduate research in cancer biology at the Duke University School of Medicine.*

That’s where Muser comes in.

Muser is a website created by Sheila Patek, a Duke biology professor who used grant money from the National Science Foundation to create a more equitable and straightforward way to connect undergraduates with professors with research opportunities. The resource allows researchers to post ongoing research positions with a direct application through the website.

Muser can sort research projects by compensation, hours, year, and project category, simplifying Duke’s incredibly complex research community by a lot.

“Muser posts research projects in 4 rounds throughout the year, a Fall round (August), a late Fall round for Spring projects (October/November), a Spring round for Summer projects (February/March), and a Spring round for Fall projects (March/April),” according to its website. Muser makes it easy to accommodate research positions into the part of your semester that works with your busy schedule.

I connected with some Duke students who have found success with the growing research platform, and though their interests were diverse, the success was all-encompassing.

“My experience with my Muser Project for the summer of 2021 was great overall,” said Elaijah Lapay, class of 2025. “It was essentially a history research assistantship helping a professor in the history department conduct research on elderly and eldercare in North Carolina. I was able to go to the NC State archives as well as archives across eastern North Carolina to really dive into the question of treatment of the elderly during the 20th century.”

Lapay’s research is so fruitful that the professor, James Chappel, the Gilhuly Family Associate Professor of History, is continuing to pursue this project for the rest of the school year. “I truly felt one-of-a-kind… I definitely feel like I’ve learned a lot and it’s sparked a passion in me for geriatrics and eldercare.”

*A look inside Dr. Laurie Sanders’s lab here at Duke University.*

“I got the chance to work in the Sanders lab under principal investigator Dr. Laurie Sanders and post-doctorate Dr.Claudia Gonzalez-Hunt!” said Shreya Goel, class of 2025. This lab was the first to link a genetic mutation to mitochondrial DNA damage which was ultimately discovered to be a marker for sporadic Parkinson’s disease.

“I get to work with human cells to induce and track mitochondrial and nuclear DNA mutations to determine their effect on the progression of the cell cycle,” Goel said. Her research position is making a difference and it allows her to gain tangible experience in a field she is passionate about.

The success stories are copious, and the opportunity that this platform has brought to prodigious students like these is without question.

At a billion-dollar research school, understanding where to begin can be intimidating. Muser alleviates these worries by connecting researchers and students through an accessible platform.

Have more questions? Visit Muser’s FAQ page to get more information and get into contact with one of Muser’s staff.

Finding the Tipping Point for Coastal Wetlands

By Robin Smith

On January 25, 2022

In Biology, Chemistry, Climate/Global Change, Environment/Sustainability, Field Research

*Cypress swamp, eastern North Carolina. Photo by Steve Anderson, Duke*

DURHAM, N.C. — The Albemarle-Pamlico Peninsula covers more than 2,000 square miles on the North Carolina coastal plain, a vast expanse of forested swamps and tea-colored creeks. Many people would probably avoid this place, whose dense thickets of cane and shrubs and waterlogged soils can slow a hike to a crawl.

“It’s hard fieldwork,” says Duke researcher Steve Anderson. “It gets really dense and scratchy. That, plus the heat and humidity mixed with the smell of sulfur and the ticks and the poison ivy; it just kind of adds up.”

But to Anderson and colleagues from Duke and North Carolina State University, these bottomlands are more than impenetrable marsh and muck and mosquitoes. They’re also a barometer of change.

*Researchers surveying plants in Alligator River National Wildlife Refuge in 2016. Photo by Mathew Stillwagon*, *North Carolina State University*

Most of the area they study lies a mere two to three feet above sea level, which exposes it to surges of ocean water — 400 times saltier than freshwater — driven inland by storms and rising seas. The salt deposits left behind when these waters recede build up year after year, until eventually they become too much for some plants to cope with.

Trudging in hip waders through stunted shrubs and rotting tree stumps, Anderson snaps a picture with his phone of a carpet of partridge berry trailing along the forest floor. In some parts of the peninsula, he says, the soils are becoming so salty that plants like these can no longer reproduce or are dying off entirely.

Along the North Carolina coast, understory plants such as this partridge berry (left) are quickly ceding ground to species such as this bigleaf marsh-elder (right) as the soils become too salty for them to thrive. Credit: Steve Anderson

In a recent study the team, led by professors Justin Wright and Emily Bernhardt of Duke, and Marcelo Ardón of NC State, surveyed some 112 understory plants in the region, making note of where they were found and how abundant they were in relation to salt levels in the soil.

The researchers identified a ‘tipping point,’ around 265 parts per million sodium, where even tiny changes in salinity can set off disproportionately large changes in the plants that live there.

Above this critical threshold, the makeup of the marsh floor suddenly shifts, as plants such as wax myrtle, swamp bay and pennywort are taken over by rushes, reeds and other plants that can better tolerate salty soils.

*Certain dwindling plants could be an early warning sign that salt is poisoning inland waters, researchers say. Credit: Steve Anderson*

The hope is that monitoring indicator species like these could help researchers spot the early warning signs of salt stress, Anderson says.

This research was supported by grants from the National Science Foundation (DEB1713435, DEB 1713502, and Coastal SEES Collaborative Research Award Grant No. 1426802).

CITATION: “Salinity Thresholds for Understory Plants in Coastal Wetlands,” Anderson, S. M., E. A. Ury, P. J. Taillie, E. A. Ungberg, C. E. Moorman, B. Poulter, M. Ardón, E. S. Bernhardt, and J. P. Wright. Plant Ecology, Nov. 24, 2021. DOI: 10.1007/s11258-021-01209-2.

*Salt is poisoning the soils past a point of no return for some marsh plants; one team is trying to pinpoint the early warning signs. By Steve Anderson.*

C is For Cookie (and for Circles of Care)

By Rebecca Williamson

On January 24, 2022

In Behavior/Psychology, Global Health, Science Communication & Education

Anyone remember the Sesame Street episode where Big Bird tackles the opioid crisis?

Me neither. However, that isn’t to say that Sesame Street isn’t doing its part to help parents and children alike to cope with this, among other pressing issues that plague our society.

Jeanette Betancourt, Ed. D. is Senior Vice President for U.S. Social Impact at Sesame Workshop, a division of the Sesame Street organization striving to positively impact children’s early learning, health, and well-being. Betancourt is deeply involved in the Sesame Street in Communities initiative (ssic.org), which she came to discuss with the Sanford School of Public Policy on January 18^th.

Sesame Street in Communities aims to bring public awareness to prominent societal issues in what Betancourt labels a “non-stigmatizing way.” Their efforts are specifically targeted to impact children coping with traumatic experiences and their families – resources on the Sesame Street in Communities website span from Elmo’s Special-Special Comfort video for children who have fallen victim to violence, to Abby’s Expressing Feelings video for children divided between divorced parents.

Not all the videos are as heavy as one might think: some of the content promotes behavioral routines, such as tooth-brushing or schedule-making, designed to build children a more stable foundation that they can use to tackle trauma, should it arise.

Some of the most recent resources posted to the *Sesame Street in Communities* website.

Betancourt says that their strategy hinges heavily upon leveling, or presenting the same messaging in a variety of mediums (videos, storybooks, live-action films), for more complete comprehension. This is reflected heavily on their website: their Autism series alone includes multiple workshops, printables, articles, videos, interactives, and storybooks. The content and learning strategies promoted by Sesame Street in Communities are all founded upon clinical research, developmental psychology, and other forms of testing to ensure that they have a measurable impact on young children and their families.

Sesame Street' Introduces 2 New Black Muppets To Teach Kids About Racial Literacy | CafeMom.com — Wes and Elijah Walker, the faces of the *Coming Together: Racial Justice* project.

One of the most recent initiatives described by Betancourt is the Coming Together: Racial Justice project. In this series of content (found on ssic.org), the viewer is introduced to the Wes and Elijah Walker, two humanoid Muppets that, according to Betancourt, are intended to represent the Black experience.

In the video, five-year old Wes and his father Elijah are sitting in the park when they are approached by Elmo, who wants to know about the pigmentation of their skin. Elijah explains to Elmo that all humans have different amounts of melanin in their skin, hence why some individuals have lighter or darker skin. Elijah also tells Elmo that, even though their skin may look different, “we’re all part of the human race.”

To make this concept easier for children to understand, Elijah connects this to the color of the changing leaves in the park, telling Elmo that leaves of different colors all came from the same tree.

If you know a child or a family that could benefit from such materials, more information can be found on ssic.org.

Post by Rebecca Williamson, Class of 2022

Blake Fauskee and the ‘Little Typos’ of Fern DNA

By Cydney Livingston

On January 21, 2022

In Biology, Genetics/Genomics, Students

Blake Fauskee, third-year Biology PhD student, initially pitched his graduate project to advisor Kathleen Pryer (Ph.D.) as an undergrad.

Fauskee, who researches RNA editing sites in ferns, told me about the project that he’s been working on for the last several years. His research could push back against the idea that DNA is the end-all, be-all molecule for encoding life as we know it.

Fauskee broke down RNA editing for me. “RNA editing is this extra step in the whole central dogma, the whole gene expression process, that happens in plant organellar DNA,” he said. This process takes place in plant mitochondrial and chloroplast genomes.

Fauskee uses a lot of metaphors to describe his work, which I find both helpful and admirable. Science can often be dense and lack feasible connections to processes that most of us are familiar with. “Basically, in [plant] DNA, there are little typos almost. The wrong nucleotide is encoded at certain spots. When those genes are going to be expressed, they get turned into RNA and then other proteins from the nucleus come in and find the little typos so that in the end you get the correct protein.”

This image shows a simplified diagram of how RNA editing works.

Fauskee calls RNA editing an “interesting and strange process” that neither animals nor humans have. His work attempts to study the evolution of this process, the patterns of RNA editing, and why it came to be. He uses DNA and RNA sequence data and the help of computational tools to do his work. He explains that when sequencing DNA, you can think of the fragmented base pairs “as little puzzle pieces.”

“So, I take all those little puzzle pieces and try to put back together the chloroplast genome, which is about 150,000 base pairs. It’s like a thousand-piece puzzle.”

Next, he figures out where the fern’s gene sequences are on the DNA strands, making use of genomic databases that contain known genomes. He then aligns RNA sequences to the genes he has mapped. Fauskee looks for the “typos” or “little differences” between the DNA and RNA: “That’s how we find the RNA editing sites.” Finally, he evaluates how the proteins would be changed by the typos in the DNA if the RNA was not edited after being transcribed.

“So, a lot of these fern genes will have a STOP codon right in the middle, which is really, really bad if you don’t fix because you are only going to get half a protein,” Fauskee said. STOP codons signal to the protein-building ribosomes that the protein is finished once it reads this portion of the RNA. Fauskee explained that these types of errors are the ones would expect organisms to lose, but it turns out they are the ones that are conserved in ferns. “Is there an extra function there? Is it helpful? Is it adaptive?” Fauskee asked.

Comparative analyses between fern species are important. By looking at whether there are common editing sites and common amino acid changes, Fauskee says, “we’re trying to understand if certain editing sites may be advantageous and what kinds of fluctuation we see between certain types of changes.”

Fauskee underscored the importance of his work. “RNA editing is a really interesting process that kind of undermines what I learned in molecular biology…They always tell you DNA is the bedrock, it’s the be-all, end-all. But what happens when the DNA is wrong? What’s the other added layer on this?”

Simply put, Fauskee, says that because of RNA editing, “We have to rethink central dogma a little bit.” In some plants, 10% of all their gene products contribute to RNA editing, Fauskee tells me. “That’s a big chunk and that’s got to be important,” Fauskee said, “Why would evolution keep such a burden going?”

Biology’s central dogma is the idea that DNA is transcribed into RNA and then translated into proteins. RNA editing adds an extra step before translation and protein production.

There may also be implications for how RNA editing sites affect the way that genetic relationships are mapped through phylogenetics. If differences between the DNA of different species at RNA editing sites, this could be misleading. Though the DNA indicates a change in base pair, RNA editing could lead to the same output in protein despite the seeming change. “If you took [RNA editing] into account,” Fauskee says, “does it give you a different answer?”

Fauskee studies ferns because of the amount of editing sites found in these plants. While flowering plants have lost editing sites over time, ferns have not. “For RNA editing, you can look at all angiosperms (flowering plants) and for the whole chloroplast genome, they might have 30-50 RNA editing sites. When you get into ferns, that number jumps up to 300-500 and I am trying to understand why.”

Botanical science first captured Fauskee’s interest while he completed his undergraduate degree in his home state at the University of Minnesota Duluth (UMD). As a sophomore at UMD, Fauskee was taken under the wing of Amanda Grusz (Ph.D.). Grusz received her PhD in biology from Duke and worked under Pryer during her own time at the university. “I’m like my advisor’s academic grandson, which is kind of funny.” Clara Howell, who is part of Fauskee’s PhD cohort and who I spoke with last Fall, is also an academic grandchild in her own lab.

Being an “academic grandson” has worked out well for Fauskee. His key advice to me for any person considering a PhD, “Make sure your advisor is not someone you just admire as a scientist, but as a person.” On a day-to-day basis, Fauskee says that advisor Katheen Pryer “is pretty hands off” but is also “one of the most supportive people ever. I’m pretty much the driver of my own ship. If I am falling off the road, she’ll push me back on the road, but she’ll give me freedom to swerve around on that road.”

Fauskee also emphasized a piece of wisdom that Pryer passed down to him. “If whatever you’ve got going on is working and everyone else is doing something different, who cares?” he said.

Though Fauskee says that “lab work can be frustrating,” getting his long analyses to run after wrangling lots of data is very rewarding. Fauskee, who does not have a background in coding or computer languages, likes to “tell people that [his] floor of biology combined is one competent coder.” When he’s not stealing bits of his biology neighbor’s code, Fauskee loves to attend Duke Basketball games and is a fan of the television show Survivor.

Post by Cydney Livingston, Class of 2022

A Conversation with Emily Levy, Soon-to-Be Biology PhD

By Cydney Livingston

On January 14, 2022

In Animals, Biology, Field Research, Students

Emily Levy studies how the physical and social environments that baboons experience affect their physiology and life outcomes. The Massachusetts native, who works under advisor Susan Alberts (PhD), is in the final year of her Biology PhD and will defend her thesis later this Spring.

Though Duke’s in-person classes have been delayed until next week, I caught up with Levy over Zoom. The wall of her home office displays a fascinating Russian map of Chicago from the cold war era that shows bridges with their weight capacity. Levy tells me that she had no idea how her husband, who is from Evanston, found the map.

Emily Levy, an almost-PhD in Duke Biology

Levy’s research stems from the Amboseli Baboon Research Project – a nearly 50-year-old, ongoing study of wild baboons in Kenya. Duke’s Alberts has been studying these baboons for over 35 years and is a renowned primatologist involved with the project. Alberts’ lab collaborates with field researchers in Kenya to receive data and samples that are imperative to much of their work.

“Something that I really appreciate about Susan and the way she runs the lab is that she starts first-year grad students on a starter project,” Levy told me. Following a discussion about Levy’s interests, this project led to her first work on dominance rank, stress levels, and what this means physiologically for baboons. Levy also “poked around” at how scientists study dominance rank and found that the methods used for assessing rank “matter a lot.”

In her more recent project, Levy is trying to figure out what early life environments mean for adulthood in baboons. “So, we know that baboons that experience a really harsh early life, if they survive to adulthood, have really, really reduced lifespans as adults – like half as long – as baboons that had no adverse events,” Levy said.

“I’m focusing on two hypotheses to get at what might be happening under the skin that could have something to do with longer term effects on health and mortality.” One hypothesis is that a tough early life environment, especially nutritional stress, could stunt baboon growth and impinge adult activities like foraging or maintaining dominance rank. The other hypothesis proposes that early life adversity disrupts immune development, leading to an immune system that is either always inflamed or produces an overpowering inflammatory response when a baboon does get sick.

The second hypothesis is one that has been supported by human research, but Levy’s preliminary results “are the exact opposite.” This highlights one facet of the importance of her research: Its implications and parallels to human health and mortality. But Levy says her research is also “cool because it’s just cool” and appreciates what her work may add to basic science beyond human application.

Images from *Scientific American’s* article about the Amboseli Baboon Research Project.

In her journey to Duke, Levy said that she “tried a few ways of studying” animals before arriving at the work she conducts now. Levy, who really liked animals, enjoyed time outside, and was “hooked on biology” in high school, began her undergraduate career at Williams College with this in mind. “In college, I took biology and neuroscience and then took animal behavior my sophomore year and was like Oooo, this is cool!,” Levy exclaimed with a big smile on her face, “And it felt sort of light-bulby.”

Along the journey to her PhD, Levy studied plants and insect pollinators and spent a few weeks in Madagascar in a tent filled with fleas. Though Levy said that these experiences of field work became “one of my favorite things about my job,” they also helped shape her trajectory as a scientist as she figured out which model systems and research questions “did and did not spark her joy.”

It was during her undergraduate thesis assessing social behavior in rats that she felt a strong “click” for studying social behavior in animals. Taking a couple years to work in a clinical research lab that conducted work on autism in humans, Levy enjoyed working on research to aid in special needs people. “But pretty quickly,” Levy said, “I was like, Alright, I don’t want to study humans for my whole life.”

“I’d basically been crossing things off my list up until this point,” Levy continued, “I now knew I wanted to study social animal behavior in non-human animals.” In her year away from research, Levy worked as an outdoor educator in Wyoming while she applied to grad school with this study plan in mind. Her time in Jackson Hole, Wyoming narrowed her interests even more, pushing her towards behavioral ecology because of her observation of an amazing, unbroken natural ecosystem.

Levy says she ultimately ended up at Duke because “the Baboon Project is amazing,” “Susan Alberts is amazing,” and “the Duke Biology Department is really wonderful.”

Images of Emily Levy conducting field work. On right, Levy is conducting field work in the Amboseli National Park in Southern Kenya.

While Levy enjoys working with Alberts and mentoring undergraduates, as well as using grant writing as a “fun way to develop really exciting ideas and hypotheses,” she also shared some of her frustrations with me. “Science is very slow — often, not always — and a project from start to finish takes a long time. And the publication process is so long. I struggle with that pace sometimes” Additionally, as someone who was raised to never take herself too seriously, Levy also said that she has felt a lot of pressure in grad school to take herself more seriously than she should as part of academic culture.

Levy loves teaching and her hope is to become a faculty member at a small liberal arts college or undergraduate institution following a post-doc, for which she is currently in the application process. Through this future work, Levy aspires to “bring undergrads through the scientific process.”

In her time away from the lab and science, Emily is an avid baker. “One of my goals in grad school has been to acknowledge and own what I am good at, and I know I am good at baking,” Levy said with a grin. Chocolate chip cookies are her specialty.

If she could give any aspiring science PhDs a word of advice, Levy offered that you should have fun and “pay attention to the non-intellectual, as well as intellectual, things that you enjoy most.” As exemplified by her path to figuring out what exactly it was in science that inspired her, Levy says not to worry as much about figuring out where you are going, and when, but reaping the lessons and insights of the experiences along the way.

*Post by Cydney Livingston, Class of 2022*

LowCostomy: the Low-Cost Colostomy Bag for Africa

By Camila Cordero

On January 11, 2022

In Biology, Biomedical Engineering, Cancer, Engineering, Environment/Sustainability, Global Health, Medicine, Students

It’s common for a Pratt engineering student like me to be surrounded by incredible individuals who work hard on their revolutionary projects. I am always in awe when I speak to my peers about their designs and processes.

So, I couldn’t help but talk to sophomore Joanna Peng about her project: LowCostomy.

Rising from the EGR101 class during her freshman year, the project is about building a low-cost colostomy bag — a device that collects excrement outside the patient after they’ve had their colon removed in surgery. Her device is intended for use in under-resourced Sub-Saharan Africa.

“The rates in colorectal cancer are rising in Africa, making this a global health issue,” Peng says. “This is a project to promote health care equality.”

The solution? Multiple plastic bags with recycled cloth and water bottles attached, and a beeswax buffer.

“We had to meet two criteria: it had to be low cost; our max being five cents. And the second criteria was that it had to be environmentally friendly. We decided to make this bag out of recycled materials,” Peng says.

For now, the team’s device has succeeded in all of their testing phases. From using their professor’s dog feces for odor testing, to running around Duke with the device wrapped around them for stability testing, the team now look forward to improving their device and testing procedures.

“We are now looking into clinical testing with the beeswax buffer to see whether or not it truly is comfortable and doesn’t cause other health problems,” Peng explains.

_{Poster with details of the team’s testing and procedures}

Peng’s group have worked long hours on their design, which didn’t go unnoticed by the National Institutes of Health (NIH). Out of the five prizes they give to university students to continue their research, the NIH awarded Peng and her peers a $15,000 prize for cancer device building. She is planning to use the money on clinical testing to take a step closer to their goal of bringing their device to Africa.

Peng shows an example of the beeswax port buffer (above). The design team of Amy Guan, Alanna Manfredini, Joanna Peng, and Darienne Rogers (L-R).

“All of us are still fiercely passionate about this project, so I’m excited,” Peng says. “There have been very few teams that have gotten this far, so we are in this no-man’s land where we are on our own.”

She and her team continue with their research in their EGR102 class, working diligently so that their ideas can become a reality and help those in need.

Post by Camila Cordero, Class of 2025

Opening the Black Box: Duke Researchers Discuss Bias in AI

By Shariar Vaez-Ghaemi

On January 6, 2022

In Artificial Intelligence, Behavior/Psychology, Data, Faculty, Policy

Artificial intelligence has not only inherited many of the strongest capabilities of the human brain, but it has also proven to use them more efficiently and effectively. Object recognition, map navigation, and speech translation are just a few of the many skills that modern AI programs have mastered, and the list will not stop growing anytime soon.

Unfortunately, AI has also magnified one of humanity’s least desirable traits: bias. In recent years, algorithms influenced by bias have often caused more problems than they sought to fix.

When Google’s image recognition AI was found to be classifying some Black people as gorillas in 2015, the only consolation for those affected was that AI is improving at a rapid pace, and thus, incidents of bias would hopefully begin to disappear. Six years later, when Facebook’s AI made virtually the exact same mistake by labeling a video of Black men as “primates,” both tech fanatics and casual observers could see a fundamental flaw in the industry.

Jacky Alciné’s tweet exposing Google’s racist AI algorithm enraged thousands in 2015.

On November 17th, 2021, two hundred Duke Alumni living in all corners of the world – from Pittsburgh to Istanbul and everywhere in between – assembled virtually to learn about the future of algorithms, AI, and bias. The webinar, which was hosted by the Duke Alumni Association’s Forever Learning Institute, gave four esteemed Duke professors a chance to discuss their view of bias in the artificial intelligence world.

Dr. Stacy Tantum, Bell-Rhodes Associate Professor of the Practice of Electrical and Computer Engineering, was the first to mention the instances of racial bias in image classification systems. According to Tantum, early facial recognition did not work well for people of darker skin tones because the underlying training data – observations that inform the model’s learning process – did not have a broad representation of all skin tones. She further echoed the importance of model transparency, noting that if an engineer treats an AI as a “black box” – or a decision-making process that does not need to be explained – then they cannot reasonably assert that the AI is unbiased.

Stacy Tantum, who has introduced case studies on ethics to students in her Intro to Machine Learning Class, echoes the importance of teaching bias in AI classrooms.

While Tantum emphasized the importance of supervision of algorithm generation, Dr. David Hoffman – Steed Family Professor of the Practice of Cybersecurity Policy at the Sanford School of Public Policy – explained the integration of algorithm explainability and privacy. He pointed to the emergence of regulatory legislation in other countries that ensure restrictions, accountability, and supervision of personal data in cybersecurity applications. Said Hoffman, “If we can’t answer the privacy question, we can’t put appropriate controls and protections in place.”

To discuss the implications of blurry privacy regulations, Dr. Manju Puri – J.B. Fuqua Professor of Finance at the Fuqua School of Business – discussed how the big data feeding modern AI algorithms impact each person’s digital footprint. Puri noted that data about a person’s phone usage patterns can be used by banks to decide whether that person should receive a loan. “People who call their mother every day tend to default less, and people who walk the same path every day tend to default less.” She contends that the biggest question is how to behave in a digital world where every action can be used against us.

Dr. Philip Napoli has observed behaviors in the digital world for several years as James R. Shepley Professor of Public Policy at the Sanford School, specifically focusing on self-reinforcing cycles of social media algorithms. He contends that Facebook’s algorithms, in particular, reward content that gets people angry, which motivates news organizations and political parties to post galvanizing content that will swoop through the feeds of millions. His work shows that AI algorithms can not only impact the behaviors of individuals, but also massive organizations.

At the end of the panel, there was one firm point of agreement between all speakers: AI is tremendously powerful. Hoffman even contended that there is a risk associated with not using artificial intelligence, which has proven to be a revolutionary tool in healthcare, finance, and security, among other fields. However, while proven to be immensely impactful, AI is not guaranteed to have a positive impact in all use cases – rather, as shown by failed image recognition platforms and racist healthcare algorithms that impacted millions of Black people, AI can be incredibly harmful.

Thus, while many in the AI community dream of a world where algorithms can be an unquestionable force for good, the underlying technology has a long way to go. What stands between the status quo and that idealistic future is not more data or more code, but less bias in data and code.

Post by Shariar Vaez-Ghaemi, Class of 2025