How do new words arise in social media?

The position of users of X (formerly Twitter) in the network predicts the fate of the new words they use, study finds

Peer-Reviewed Publication

PLOS

 

image: 

Fig 2. Two trajectories of lexical innovations.

The usage rate per month of two changes (left) and two buzzes (right) represented by a rolling average with a three-month window (blue), as well as the result of the curve fitting (green). The three diffusion phases are represented by the grey shading in the background

view more 

Credit: Tarrade et al., 2024, PLOS Complex Systems CC-BY 4.0 (https://creativecommons.org/licenses/by/4.0/)

The more centrally connected someone is within their social media network, the more likely that new words they use will become adopted into mainstream language, according to a new study published this week in PLOS Complex Systems by Louise Tarrade of École Normale Supérieure, France, and colleagues.

Language evolves within a social context and variations in a language are always in competition with each other. In everyday language, words are constantly being created, but not all these words persist.

In the new study, researchers analyzed more than 650 million tweets written in French between 2012 and 2014 to identify 400 words that were new to appear in the social media network X (then called Twitter) during that time. Then, they tracked the diffusion of these words over the following five years, and looked at the position and connectivity of users who adopted the words.

The team found that initial adopters of new words, or “lexical innovations” are similar to each other, regardless of whether the innovations later succeed (as language changes) or fade away (as temporary buzzes). However, in the later propagation phase for lexical innovation, there were statistically significant differences between changes and buzzes. On average, the words that eventually persisted were used by people who were more central to their community, and remained in circulation at low levels for a longer period before entering a growth phase (18.5 months in circulation compared to 6.5 months for buzzes). Words that became only temporary buzzes were used by people with less central positions within a social network and had a more rapid rise in use – followed by a rapid decline.

“Our research examines lexical innovations at scale across millions of social media users,” the authors say. “We show that words adopted by users who are more central in their community and easily in contact with other communities become established in the language, and vice versa. Thus, the position in the network of speakers who adopt these words is enough to predict their fate.”

The authors add: “Doing sociolinguistics with digital data and computational methods offers the opportunity to scale up and reveal social dynamics at the population level.”

############

In your coverage, please use this URL to provide access to the freely available article in PLOS Complex Systems: https://journals.plos.org/complexsystems/article?id=10.1371/journal.pcsy.0000005

Citation: Tarrade L, Chevrot J-P, Magué J-P (2024) How position in the network determines the fate of lexical innovations on Twitter. PLOS Complex Syst 1(1): e0000005. https://doi.org/10.1371/journal.pcsy.0000005

Author Countries: France

Funding: J.-P. M., J.-P. C. and L.T. are grateful to the ASLAN project (ANR-10-LABX-0081, https://aslan.universite-lyon.fr/) of the Université de Lyon for its financial support within the French program "Investments for the Future" operated by the National Research Agency (ANR). The data collection has been supported by the SoSweet ANR project (ANR-15-CE38-0011-03, https://anr.fr/) attributed to J.-P. M. and J.-P. C. The authors are also grateful to University of Grenoble Alpes and Ecole Normale Supérieure de Lyon for the support for publication. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

---

Journal

PLOS Complex Systems

DOI

10.1371/journal.pcsy.0000005 

Method of Research

Observational study

Subject of Research

People

 

In US community efforts to prevent childhood obesity, women leaders with more experience and connections have the greatest impact

PLOS

 

image: 

Fig 3. Cross-section of coalition-committee networks by wave of data collection.

This figure, panels A-L, illustrates the cross-section (by wave of data collection) network position movement of two coalition-committee members in each community who demonstrated the highest increases in their Knowledge and Engagement scores; it combines our positional, triad, and community detection analysis with changes in two of our key outcomes. Circles are coalition-committee members, with the two largest circles indicating the coalition-committee members who saw the largest increase in their Knowledge and Engagement scores. Triangles are those with whom coalition-committee members discuss childhood obesity prevention, termed first-degree alters. The colors correspond to community clusters that were identified using the SPLA algorithm.

view more 

Credit: Moore et al, 2024, PLOS Complex Systems, CC BY 4.0

In US community efforts to prevent childhood obesity, women leaders with more experience and connections have the greatest impact.

####

In your coverage, please use this URL to provide access to the freely available article in PLOS Complex Systems: https://journals.plos.org/complexsystems/article?id=10.1371/journal.pcsy.0000004  

Article Title: Determinants and facilitators of community coalition diffusion of prevention efforts

Author Countries: United States

Funding: TM, MP, and CE received funding from the JPB Foundation (PR0580) https://www.jpbfoundation.org. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

---

Journal

PLOS Complex Systems

DOI

10.1371/journal.pcsy.0000004 

COI Statement

Competing interests: The authors have declared that no competing interests exist.

SOCIAL ECOLOGY

Re-creations of 1870s railway photos reveal profound change to Kansas, Colorado plains

University of Kansas

 

image: 

A fascinating new book by University of Kansas researcher Town Peterson chronicles transformation on the plains of Kansas and western Colorado using repeat photography — contemporary re-creations of 1870s photos — to reveal startling changes to the landscape.

view more 

Credit: Town Peterson

LAWRENCE — A fascinating new book chronicling transformation on the plains of Kansas and western Colorado uses repeat photography — contemporary re-creations of 1870s photos — to reveal startling changes to the landscape.

Its author isn’t just a photographer and veteran of years of “Kansas-ing” — his term for searching off-the-beaten-path curiosities across the Sunflower State — but also a University Distinguished Professor of Ecology & Evolutionary Biology at the University of Kansas and senior curator of ornithology at KU’s Biodiversity Institute and Natural History Museum.  

While Town Peterson usually focuses his research on the geography of biodiversity, tropical ornithology and systematics, distributional ecology and disease-transmission risk mapping, his new book, “One Hundred and Fifty Years of Change on the Great Plains,” is a bit different. 

“I've gotten interested in what you could call ‘historical ecology,’ essentially going deeper in time than what we can sample and study now,” Peterson said. “This historical photograph series from 1873 has become really interesting. First, it’s unique because the photographer was hired to travel the length of the Kansas Pacific railroad and take photos, apparently for promotional purposes. At nearly every major stop along the railroad, Robert Benecke took photographs. By nothing short of a miracle, most or all of these photos survived and ended up in the collections of Southern Methodist University.”

Other repeat-photography projects have re-created the work of 19th century photographers in the region, but none with the landscape locations as precisely described as Benecke’s.

“There's also an older collection from a man named Alexander Gardner, who also traveled the length of the Kansas Pacific railroad, but when it was being surveyed around 1867-69,” Peterson said. “Benecke took his photos in 1873, when the railroad was mostly in place and functioning. The Gardner photographs have been part of a repeat-photography effort about 20 years ago, which was fascinating. But Benecke’s material had never been touched, even though Benecke’s photo locations were better described. Gardner would say things like ‘334 miles west of St. Louis,’ while Benecke would reference specific towns like Brookville or Russell.”

During travel restrictions put in place during the COVID-19 pandemic, Peterson found himself unable to conduct his usual international research and education. So he decided to focus on a Kansas-based project and started working with Benecke’s material, distilling it down to landscape photographs that he could find and photograph today.

“It came to around 50 landscape views,” he said. “Over the last couple of years, I traveled back and forth between Kansas City and Denver a few times. Some were painfully hard to relocate, and others, even when you knew where they were, were painfully hard to photograph… brambles, ticks, trees, etc.”

Even when Peterson was able to pinpoint sites of Benecke’s photos, sometimes changes to the landscape made exact re-creation through the camera viewfinder impossible. Indeed, the KU scientist had to make very difficult choices in how to best re-create Benecke’s original images. 

“There were some tough decisions involved,” Peterson said. “Do I take a picture of an ugly cement drainage culvert, or do I move a half-mile down the river and photograph the mouth of the same river where it now empties into the Kansas River? Do I take a picture of a dense forest that wasn’t there in the 19th century, or do I use a drone to get a shot of the broader landscape — something closer to what the original photographer might have taken?”

The changes over a century and a half, revealed by comparing Benecke’s and Peterson’s images side by side, are striking. Peterson said that, as an evolutionary biologist, the starkest difference was the presence of so many trees today in places that had none 150 years ago.

“We live in the Great Plains, yet if you look out pretty much any window here in Lawrence, it doesn't look like plains — it looks like a forest,” he said. “I've been here for 30 years, and it has always pretty much looked like this. So, I got interested in how long-term that view is. I've come to appreciate how our Great Plains is very different from the Great Plains that was settled 150 years ago by Europeans, much less farther back still. You get west of Manhattan, and there’s pretty much not a single tree in any of Benecke’s photographs. Now, there are trees all the way to Denver, at least in the towns and cities.”

Peterson said this “afforestation” — the opposite of deforestation — is a pervasive force across the Great Plains. Indeed, the KU researcher is carrying out studies to explore the phenomenon more.

“That, to me, was the biggest lesson on the science side,” he said. “I’m involved in studies where we’re exploring how much the biodiversity has changed at sites along this route. We’re seeing how dozens of species have taken advantage of the now-forested landscape to expand their ranges westward into the Great Plains.”

Some hallmarks of the 19th century American West are notably absent from Benecke’s photos, such as American bison or signs of Native American life and culture. Peterson thinks these omissions spring from the promotional purposes of the Benecke photos, which were in part intended to sell railroad-owned land at a massive profit.

“It was all about getting settlers of European descent from the East to come out and settle on this landscape,” he said. “If they could sell the land deeded to the railroad by the U.S. government, it was pure profit. Life in western Kansas wasn’t at all easy, but the settlers didn’t need to know that until they had already paid their money.”

“One Hundred and Fifty Years of Change on the Great Plains” is a book that is intended to be accessed freely and openly. It is available as an e-book via KU ScholarWorks. A hardcover version is available for purchase at the cost of printing and shipping.

 

New species of pāua found

University of Otago

The naming of a new species of pāua further highlights the importance of biodiversity research in Aotearoa.

Described in a study led by the University of Otago – Ōtākou Whakaihu Waka and the Museum of New Zealand Te Papa Tongarewa, the species, Haliotis pirimoana, is only found at Manawatāwhi Three Kings Islands, off the northern North Island.

Lead author Kerry Walton, an invertebrate curator at Te Papa who is undertaking his PhD in the Department of Zoology, says this is one of many species that are only found on Manawatāwhi.

“We are facing a biodiversity crisis. Species around the world are going extinct faster than we can discover and describe them,” he says.

“The Manawatāwhi pāua may not be facing any short-term threats to its survival but it is another indicator of how unique and important that region of Aotearoa is.”

Haliotis pirimoana resembles other small pāua from the North Island but the shell has finer sculpture across its back and a subtly different shape.

Published in the international peer-reviewed journal Molluscan Research, the study aimed to understand how many species of small pāua there are in the country.

To do this, the research team, including the Otago Department of Zoology’s Associate Professor Nic Rawlence and Professor Hamish Spencer, and Dr Bruce Marshall from Te Papa, applied new “ancient DNA” methods to shells from the collections at Te Papa and NIWA.

“These new methods allow short and damaged fragments of DNA to be extracted and sequenced with far greater efficiency than previous methods, unlocking the genetic secrets trapped in these precious specimens,” Associate Professor Rawlence says.

People can now more confidently discuss and interpret relationships between the different forms and populations of small pāua in Aotearoa, he says.

Professor Spencer says that biodiversity is critical, and effective resource and conservation management relies on having a robust underlying understanding of it.

“We need to know what species are restricted to a region and how they might respond to change. This allows decision-makers to identify vulnerabilities and best manage resourcing,” he says.

Mr Walton says while most species will not yield commercial applications, such as the next big pharmaceutical, people depend on biodiversity for new materials and medicines.

“Research in these areas is underpinned by biodiversity research, and it is our responsibility to protect biodiversity for future generations.”

While there is a decent level of understanding of biodiversity in Aotearoa, new species are still being discovered and more than a third of our known marine mollusc species are yet to be named.

“With pāua being the cultural icon that they are, we hope this work will raise a few eyebrows and encourage people to look more closely the next time they are at the coast.”

---

Journal

Molluscan Research

DOI

10.1080/13235818.2024.2390476 

Method of Research

Case study

Subject of Research

Animals

Article Title

Haliotis virginea Gmelin, 1791 and a new abalone from Aotearoa New Zealand (Mollusca: Gastropoda: Haliotidae)

GIFT ECONOMY

Explaining the mechanism of social evolution driven by gift giving

A study explores competitive gift giving and the emergence of economic and social disparities

PLOS

 

image: 

Schematic of gift-giving interaction. The donor endows a gift to the recipient. If the recipient reciprocates by adding interest within a certain time limit, the donor and recipient have an equal friendship. However, if the recipient cannot reciprocate appropriately, the recipient becomes subordinate to the donor.

view more 

Credit: Kaneko and Itao, CC-BY 4.0, https://creativecommons.org/licenses/by/4.0/)

New findings provide quantitative criteria for classifying social organizations in human history, together with potential explanatory variables that can be empirically measured for anthropology, history, and archaeology, according to a study published September 3, 2024 in the open-access journal PLOS Complex Systems by Kenji Itao and Kunihiko Kaneko from the University of Tokyo, Japan and Copenhagen University, Denmark (Kaneko) and the RIKEN Center for Brain Science, Japan (Itao).

Human societies have experienced transitions between different types of organizations, including bands, tribes, chiefdoms, and kingdoms. However, quantitative characterizations of the types and mechanisms of transitions have yet to be established. Meanwhile, anthropologists have observed that gift giving is prevalent in traditional societies and that it enhances social status by imposing reciprocal obligations on others. 

In the new study, Itao and Kaneko present a simple model of competitive gift giving that describes how gifts bring goods to the recipient and honor to the donor, and simulate social change. In particular, the authors demonstrated that competitive gift giving generates socioeconomic disparities characterized by power law distributions of wealth and social reputation. 

Numerical simulations revealed transitions between four phases with different distribution shapes of wealth and social reputation. The phases included the band, without economic or social disparities; the tribe, with economic but without social disparities; the chiefdom, with both; and the kingdom, with economic disparity and weak social disparity except for an outlier, namely, the monarch. The emergence of strong disparities was characterized by power law distributions and was attributed to the rich-get-richer process. In contrast, the absence of such a process led to exponential distributions due to random fluctuations. In addition, the phases depended on parameters characterizing the frequency and scale of gift interactions.

In other words, the emergence of several phases of social organizations are quantitatively characterized by the shape of the wealth and reputation score distributions. In the band phase, both distributions are exponential. In the tribe phase, only the wealth distribution obeys a power law. In the chiefdom phase, both distributions are power laws. And in the kingdom phase, the score distribution is exponential for all individuals except the monarch.

Overall, the results demonstrate the emergence of four phases of social organizations characterized by the degrees of economic and social disparities. The study also provides theoretical support for empirical findings that band societies have less economic inequality than do the other classes, that the presence of social inequality distinguishes chiefdoms and kingdoms from the others, and that monarchs are outliers in the distribution. According to the authors, their constructive model, guided by social scientific theory, can provide the basic mechanistic explanation of social evolution and integrate theories of the social sciences.

The authors add: "Gift-giving interactions drive social change. Its frequency and extent determine social organization structures."

############

In your coverage, please use this URL to provide access to the freely available article in PLOS Complex Systems: https://journals.plos.org/complexsystems/article?id=10.1371/journal.pcsy.0000001   

Citation: Itao K, Kaneko K (2024) Emergence of economic and social disparities through competitive gift-giving. PLOS Complex Syst 1(1): e0000001. https://doi.org/10.1371/journal.pcsy.0000001

Author Countries: Denmark, Germany, Japan

Funding: This research was supported by the Grant-in-Aid for Scientific Research (A) (20H00123) from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT) of Japan, JSPS KAKENHI Grant Number JP21J21565 (KI), and Novo Nordisk Fonden Grant Number NNF21OC0065542 (KK). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

---

Journal

PLOS Complex Systems

DOI

10.1371/journal.pcsy.0000001 

COI Statement

Competing Interests: The authors have declared that no competing interests exist.

 

Kinsey Institute distinguished research scientist wins ISPNE 2024 Bruce Mcewen Lifetime Achievement Award for groundbreaking research on oxytocin and social behavior

Kinsey Institute

 

image: 

Kinsey Institute Distinguished Research Scientist Dr. C. Sue Carter

view more 

Credit: Kinsey Institute at Indiana University

Dr. C. Sue Carter, Distinguished Research Scientist and Director Emerita of the Kinsey Institute, has been awarded the 2024 Bruce McEwen Lifetime Achievement Award by the International Society of Psychoneuroendocrinology. This prestigious award recognizes Dr. Carter’s lifetime of pioneering research that has significantly advanced our understanding of the biological and social underpinnings of human behavior.

A career biologist, Dr. Carter specializes in the endocrinology of love and social bonds. Dr. Carter is renowned for her discovery of the critical role of oxytocin in social bonding, stress management, and emotional regulation. She was the first person to detect and define the physiology of monogamy through her research on the prairie vole, helping lay the foundation for the studies of behavioral and developmental effects of oxytocin and vasopressin in humans.

Dr. Carter's research has had a profound impact on multiple scientific fields including neuroscience, psychology, and endocrinology, helping to illuminate the complex relationship between hormones, social behavior, and mental health. She continues to collaborate widely, contributing to our knowledge of the neurobiology of social behaviors and the concept of "sociostasis," which is the capacity to use sociality to anticipate and cope with challenges across the lifespan.

The International Society of Psychoneuroendocrinology has awarded the prestigious Lifetime Achievement Awards for over 20 years to a distinguished line of outstanding scientists in the field of psychoneuroendocrinology, for their contributions to our understanding of brain-body interactions.

About the Kinsey Institute at Indiana University

For almost 80 years, the Kinsey Institute has been the premier research institute on human sexuality and relationships and a trusted source for evidence-based information on critical issues in sexuality, gender, reproduction, and well-being. The Kinsey Institute's research program integrates scholarly fields including neuroscience, psychology, public health, anthropology, history, and gender studies. Kinsey Institute outreach includes traveling art exhibitions, public scholarships, research lectures, and a human sexuality education program. Visit our website kinseyinstitute.org and follow us on LinkedIn.

Study at Pennington Biomedical Research Center to evaluate THC, CBD benefits for dementia-related agitation

Pennington Biomedical currently recruiting hospice-eligible dementia patients “LiBBY” research study

Business Announcement

Pennington Biomedical Research Center

Pennington Biomedical Research Center’s Dr. Jeff Keller is evaluating the potential for delta-9-tetrahydrocannabinol, or THC, and cannabidiol, or CBD, to reduce the behaviors indicating agitation, distress or anxiety in patients with Alzheimer’s disease or other forms of dementia. The study is designed for hospice-eligible patients who are either receiving hospice care or who are eligible for hospice, and who are exhibiting agitation concurrently with a diagnosis of dementia. There are currently no FDA-approved medications to treat agitation at the end-of-life stages in dementia patients. 

The “Life’s End Benefits of Cannabidiol and Tetrahydrocannabinol” or LiBBY, study is the first-of-its-kind national study funded by the National Institute of Aging. The placebo-controlled, double-blinded study is designed to help researchers further explore these compounds for their potential to ease the burden of patients and caregivers.  

“The restlessness, anxiety, and distress that frequently accompanies Alzheimer’s and dementia can be frustrating and frightening for both the patient and their caregivers,” said Dr. Keller, who is the director of Institute for Dementia Research & Prevention at Pennington Biomedical and a professor of aging and neurodegeneration. “The LiBBY study intends to explore whether THC and CBD can be safe and effective options for managing agitation in dementia patients. This landmark, multisite study will increase the pace of THC clinical research nationwide, and Pennington Biomedical’s participation makes it much easier to initiate THC research at this facility in the future.” 

Upon joining the study, participants will complete a baseline visit and will be randomly assigned to one of two study groups: one receiving the active study drug, and one receiving a placebo. The active study will last for 12 weeks and will include one screening visit, six in-person visits, and five evaluations by phone.  

Before the study concludes, all participants will have the opportunity to request to continue their participation in the “Open Label Extension” portion of the study. Though participants will not know if they are receiving the study drug or the placebo during the study, all participants will receive the active study drug in the “Open Label Extension” period. 

“Dementia, especially when coupled with agitation, can be frightening for patients and heartbreaking for caregivers, who frequently face burnout. As other studies have shown THC can help in treating post-chemotherapy nausea, the LiBBY study is a practical next step in exploring the benefits of these compounds and offering hope to those facing this difficult challenge,” said Dr. John Kirwan, Executive Director of Pennington Biomedical. “Dr. Keller and his colleagues have worked incredibly diligently to bring this study to Pennington Biomedical. We are proud to have this study here and to have it position the Center for similar studies in the future.” 

Pennington Biomedical is one of 11 research institutions to facilitate the LiBBY study and exploring the use of THC for dementia-related agitation. Pennington Biomedical, Georgetown University Medical Center, the University of Kentucky, and the Ralph H. Johnson VA Medical Center are currently recruiting participants.

The research facility is working with hospice care providers in Capital Region and beyond to share the details of this study with patients and their families. Carpenter Health network, a large Baton Rouge-based hospice group, is one of the care providers to help identify patients and caretakers who may be eligible for the study.

"The Carpenter Health Network is proud to join with other late-stage care facilities and networks to share this study with the families we work with daily," said Raoul Manalac, Medical Director at St. Joseph Hospice – Baton Rouge. "Dementia is accompanied by many challenges, with agitation and anxiety among the most frightening and frustrating for patients and families, and there are few options available to provide relief. This innovative study is driven by both the quest for new knowledge, as well as compassion for families dealing with dementia.”  

Hospice groups and end-of-life care providers in South Louisiana will be instrumental connecting eligible patients with the study. Those providers or groups who are interested can contact Dr. Jeff Keller at jeffrey.keller@pbrc.edu. 

About the Pennington Biomedical Research Center

The Pennington Biomedical Research Center is at the forefront of medical discovery as it relates to understanding the triggers of obesity, diabetes, cardiovascular disease, cancer and dementia. The Center conducts basic, clinical, and population research, and is a campus of the LSU System. The research enterprise at Pennington Biomedical includes over 530 employees within a network of 44 clinics and research laboratories, and 13 highly specialized core service facilities. Its scientists and physician/scientists are supported by research trainees, lab technicians, nurses, dietitians, and other support personnel. Pennington Biomedical is a state-of-the-art research facility on a 222-acre campus in Baton Rouge. For more information, see www.pbrc.edu.

 

Illinois scientists to test modernized genetic model for optimized crop breeding

University of Illinois College of Agricultural, Consumer and Environmental Sciences

 

image: 

The University of Illinois Urbana-Champaign's Alex Lipka (left) and Colorado State University's Geoffrey Morris (right) will test the omnigenic model for its utility in crop breeding, thanks to new funding from the National Science Foundation. 

view more 

Credit: University of Illinois Urbana-Champaign and Colorado State University

URBANA, Ill. — The National Science Foundation (NSF) has funded University of Illinois Urbana-Champaign research that aims to connect the dots between quantitative and molecular genetics and improve crop breeding.

The four-year, $795,000 grant investigates new theories on how genetics influence complex crop traits, such as yield or grain quality. These traits are controlled by lots of different genes — sometimes hundreds or thousands — which makes untangling their contributions difficult. Crop breeders use a host of advanced genetic tools to predict and select desirable complex traits, but these tools rely on outdated genetic understanding, believes project leader Alex Lipka.

“The theory used to quantify genetic contributions to traits in statistical models stems back from 1918. In 1918, they didn’t have the central dogma of molecular biology, so they didn’t even know that DNA had two strands. There are over a century of advancements that have not been incorporated into the most widely used models to quantify genetic architecture,” said Lipka, an associate professor in the Department of Crop Sciences, part of the College of Agricultural, Consumer and Environmental Sciences (ACES) at Illinois. 

An emerging genetic framework called the omnigenic model incorporates modern advances in molecular biology into classical genetic theory. The omnigenic model divides all the genes in an organism’s genome into two components: core genes and peripheral genes. According to the model, the core genes directly control the trait of interest. If, for example, several core genes for plant height are switched on, the plant should be taller. 

Peripheral genes, on the other hand, do not directly control the trait but can still subtly impact it. These genes are involved in cellular processes that influence how the core genes direct the trait. For example, a peripheral gene might produce a protein that can travel within the cell and change the activity of a plant height core gene. While the effects of the peripheral genes may be small individually, added together they can contribute even more to genetic variability than core genes. 

If the omnigenic model is correct, Lipka believes that incorporating peripheral genes could advance breeding tools. “If we can harness the collective effects of the peripheral genes, then there can be really powerful ramifications for getting and selecting for optimal trait values,” Lipka said. 

Lipka and his collaborator Geoffrey Morris at Colorado State University, who also received an NSF grant for this project, will develop statistical methods for testing the omnigenic model in crops. 

“We don’t currently have the statistical tools to properly assess evidence of the omnigenic model,” Lipka said. “We’re going to develop these tools and test them out in a biologically rigorous manner.”

They plan to use a software package previously developed by Lipka’s team to simulate how core genes, peripheral genes, and the interaction between genes could affect complex traits in a simulated crop population. Their simulations will be informed with data from Arabidopsis, a model plant species, and sorghum, a climate-resilient crop widely eaten in areas of the world with food insecurity. 

"It's difficult for plant breeders to keep pace with a changing climate and increasing food demand," said Morris, whose team supports plant breeding programs around the world. "In this project, new methods will first be rigorously tested with data sets from ongoing breeding partnerships in the U.S., Senegal, and Haiti. Ultimately, though, our goal is to see these methods deployed by plant breeders to identify high-yielding, climate-resilient varieties."

They will simulate multiple populations with different types of selection and selection intensity, repeat this for several generations, and ultimately quantify evidence for or against the omnigenic model. By the end of the project, they will put all of their work into a new software package that other researchers can use to test and apply the model.

“In some of the preliminary studies, the omnigenic model actually seems like it might be working, which is just really cool,” Lipka said. 

Lipka is also affiliated with the College of Liberal Arts and Sciences, the Carl R. Woese Institute for Genomic Biology, and the National Center for Supercomputing Applications at Illinois.

New machine learning model offers simple solution to predicting crop yield

Feature engineering improves upon genotype-by-environmental interaction model

University of Arkansas System Division of Agriculture

 

image: 

Sam Fernandes, left, assistant professor of agricultural statistics and quantitative genetics with the Arkansas Agricultural Experiment Station, and Igor Fernandes, statistics and analytics master's student, have worked to improve a crop yield prediction model using environmental data.

view more 

Credit: U of A System Division of Agriculture photo by Paden Johnson

By John Lovett

U of A System Division of Agriculture

FAYETTEVILLE, Ark. — A new machine-learning model for predicting crop yield using environmental data and genetic information can be used to develop new, higher-performing crop varieties.

Igor Fernandes, a statistics and analytics master’s student at the University of Arkansas, entered agriculture studies with a data science background and some exposure to agronomy as an undergraduate assistant for Embrapa, the Brazilian Agricultural Research Corporation. With an outsider’s perspective and a history working with environmental data through one of his former advisers, he developed a novel approach to forecasting how crop varieties will perform in the field.

His interest in the subject led to a recently published study co-authored with his adviser, Sam Fernandes, an assistant professor of agricultural statistics and quantitative genetics with the Arkansas Agricultural Experiment Station, the research arm of the University of Arkansas System Division of Agriculture.

The study, published in the Theoretical and Applied Genetics journal, is titled “Using machine learning to combine genetic and environmental data for maize grain yield predictions across multi-environment trials.”

“Igor came in from statistics with no genetics background,” Sam Fernandes said. “So, he had this idea that was not at all what we would use in genetics, and it was just surprising that it worked well.”

Igor Fernandes’ model, which focused on environmental data, led him to a close second in this year’s international Genome to Fields competition. Co-authors of the study that stemmed from the competition entry included Caio Vieira, an assistant professor of soybean breeding for the experiment station, and Kaio Dias, assistant professor in the department of general biology at the Federal University of Viçosa in Brazil.

Environment and genetics

While the competition entry showed environmental data alone worked better than expected at predicting crop yield, the researchers saw an opportunity to build a comprehensive study that compared the novel approach to established prediction models used in genomic breeding.

Genomic breeding, a process of screening thousands of candidates for field trials based on DNA alone, can save time and resources needed to develop a new plant variety, such as growing better in drought conditions. An important part of genomic breeding involves genomic prediction to estimate a plant’s yield using its DNA.

“Let’s say you have thousands of candidates, and you get the DNA from all of them,” Sam Fernandes explains. “Based on the DNA along with information from previous field trials, you are able to tell which one will be the highest yielding without planting it in the field. So, you’re saving resources that way. This is genomic prediction.”

Adding information into a model on how that plant would interact with environmental conditions increases the accuracy of the genomic prediction and is becoming more common as more environmental data from testing centers becomes available. The practice is called “enviromics.” Still, there is no consensus on the best machine learning approach to combine environmental and genetic data.

“One advantage of including the environment information in the models is that you can address what we call genotype-by-environmental interaction,” Sam Fernandes said. “Since the environment does not affect all of the individuals in the same way, we try to account for all of that, so we are able to select the best individual. And the best individual can be different depending on the place and season.”

The study used the same data on corn plots from the Genomes to Fields Initiative that were used in the competition, but the researchers adjusted inputs as genetic, environmental, or a combination of both in “additive” and “multiplicative” manners. When including environmental and genetic data in a more straightforward “additive” manner, the prediction accuracy was better than the more complicated “multiplicative” manner.

The simpler model took less time for the computer to process, and the mean prediction accuracy improved 7 percent over the established model. The experiment was validated in three scenarios typically encountered in plant breeding.

“One of the unique things that Igor did is how he processed the environmental data,” Sam Fernandes said. “There are fancier models that people can throw in all sorts of information. But what Igor did is a simple, yet efficient way of combining the genetic and environmental data using feature engineering to process the information and get a summary of variables that is more informative.”

Collectively, the researchers say the results are promising, especially with the increasing interest in combining environmental features and genetic data for prediction purposes. Their immediate goal is to apply it to increase the capability of screening genotypes for field trials.

​To learn more about Division of Agriculture research, visit the Arkansas Agricultural Experiment Station website. Follow on 𝕏 at @ArkAgResearch. To learn more about the Division of Agriculture, visit uada.edu. Follow us on 𝕏 at @AgInArk. To learn about extension programs in Arkansas, contact your local Cooperative Extension Service agent or visit uaex.uada.edu.

About the Division of Agriculture

The University of Arkansas System Division of Agriculture’s mission is to strengthen agriculture, communities, and families by connecting trusted research to the adoption of best practices. Through the Agricultural Experiment Station and the Cooperative Extension Service, the Division of Agriculture conducts research and extension work within the nation’s historic land grant education system.

The Division of Agriculture is one of 20 entities within the University of Arkansas System. It has offices in all 75 counties in Arkansas and faculty on five system campuses.

The University of Arkansas System Division of Agriculture offers all its Extension and Research programs and services without regard to race, color, sex, gender identity, sexual orientation, national origin, religion, age, disability, marital or veteran status, genetic information, or any other legally protected status, and is an Affirmative Action/Equal Opportunity Employer.

---

Journal

Theoretical and Applied Genetics

DOI

10.1007/s00122-024-04687-w 

Method of Research

Data/statistical analysis

Subject of Research

Not applicable

Article Title

Using machine learning to combine genetic and environmental data for maize grain yield predictions across multi‑environment trials