MicroRNAs are small molecules that can guide proteins to decrease gene expression, and engineering artificial versions allows scientists to target specific genes for crop improvement.

“Though these microRNA molecules are very tiny, their impacts are huge,” said Xiuren Zhang, Ph.D., Christine Richardson Endowed Professor in the Texas A&M College of Agriculture and Life Sciences Department of Biochemistry and Biophysics, adjunct professor in the Texas A&M College of Arts and Sciences Department of Biology, and principal investigator of the study.

Changhao Li, Ph.D., and Xingxing Yan served as co-first authors of the study, with supervision from Xiuren Zhang, Ph.D. The team’s work has substantially revised the current understanding of microRNA biogenesis in the model organism Arabidopsis thaliana. (Jiaying Zhu/Texas A&M AgriLife)

Using precise mutations and a clever experimental design, Texas A&M AgriLife researchers reevaluated the landscape of microRNAs in the model organism Arabidopsis thaliana and found that fewer than half of them were correctly identified as microRNAs, while the others are miscategorized or require further investigation.

In addition to clarifying genuine microRNA molecules in Arabidopsis thaliana, the study supplies an effective experimental design for repeating the analysis in other crops and even in animals, which likely need a similar review. The team’s discoveries also helped them create updated guidelines for designing artificial microRNAs, opening the door to improvement in crops like corn, wheat, soybeans and rice.

Xingxing Yan, a graduate research assistant, and Changhao Li, Ph.D., a postdoctoral research associate, were co-first authors of the study. It was funded by the National Institutes of Health, National Science Foundation and the Welch Foundation.

A decade-old endeavor

MicroRNAs have a uniform length of around 21 to 24 nucleotides. But in plants, Zhang said their precursors come in a range of shapes and sizes.

Because of the precursors’ structural diversity, determining which key features are most important for their processing has been a challenge, and it’s left the question of how microRNAs are generated in plants largely unexplored and unverified.

Arabidopsis thaliana, also known as thale cress and mouse-ear cress, is a model organism for plant biology. Its relatively small genome, quick growth and production of many seeds make it exceptionally useful in research. (Xingxing Yan/Texas A&M AgriLife)

About 10 years ago, Zhang said, he and his lab found a pattern between a loop on the precursor microRNA structure and the first cut site. This initial cut is significant because it determines the first nucleotide on the mature microRNA molecule, an important factor for directing it to the correct location in a cell.

Unfortunately, of the 326 posited microRNA precursors in Arabidopsis thaliana, only a few had the ideal reference loop that Zhang’s lab found — according to the computational models, at least.

“The models are based on pure chemistry,” Zhang said. “They focus only on the free energy, on what should be the most stable form. But it couldn’t explain why so many diverse precursors can end up with products of the same size.”

Rather than relying on the models, Zhang’s lab sought to verify the microRNA precursors within plants. They wanted to find the first cut sites on the precursors and confirm their structural determinants within cells.

Unexpected findings

To do this, the researchers made highly specific mutations to the dicer protein, which, as its name implies, is responsible for making precise cuts to the microRNA precursor. Normally, the protein acts like two hands that hold a double strand of precursor RNA and cut at a site in each strand concurrently before releasing the RNA molecule.

“We made point mutations at two locations separately in the dicer-like protein to make them semi-active,” Yan said. “That way, they can only cut one strand and stop before further processing. This gives us a chance to capture the intermediate products of the microRNA precursor, telling us the initial processing sites and that first nucleotide.”

Their results showed that only 147 of the 326 posited microRNA precursors interact with the dicer protein definitively, marking these as genuine microRNA precursors. Eighty-one didn’t interact at all, suggesting they should be reclassified as a different type of RNA. Around 100 require further investigation.

The team also used an advanced high-throughput technique and new computational method to map out the structures of microRNA precursors in their natural cell conditions and found that, of the 147 genuine microRNA molecules, about 95% of their structures in cells differed from computer predictions.

“We found several results quite different from predictions and from the literature,” Li said. “We were able to combine biochemical results with next-generation sequencing to get more information, and now our understanding of the structures is much more accurate.”

The future

The team still has more microRNA precursors to validate in Arabidopsis thaliana, but Zhang said they are excited to pursue collaborations to investigate microRNA processing in agricultural crops for more practical applications.

“We want to find out more about what kind of microRNAs are in other crops, how they’re processed and how we can make artificial microRNAs in them,” he said. “This study provides resources that can be used widely, and now we can use it to revisit other crops, find what needs to be corrected, and see what else we can do with this tool.”

The study, led by Pankaj Arora, M.D., and Naman Shetty, M.D., examined data from 35,206 participants in the UK Biobank. The researchers investigated the clinical correlates of TTR levels, differences in TTR levels based on genetic variations and the association of TTR levels with health outcomes.

Arora and his team found that lower TTR levels are significantly associated with an increased risk of heart failure and all-cause mortality. Specifically, individuals with low TTR levels had a 17 percent higher risk of heart failure and an 18 percent higher risk of death from any cause compared to those with higher TTR levels. These findings were even more pronounced in individuals carrying the V142I TTR gene variant, which is known to destabilize the TTR protein.

The study revealed that TTR levels were lower in females compared to males and were influenced by several health factors. Higher systolic and diastolic blood pressure, total cholesterol, albumin levels, triglyceride levels, and creatinine levels were associated with increased TTR levels. Higher C-reactive protein levels were linked to lower TTR levels. Notably, carriers of the V142I TTR gene variant had significantly lower TTR levels compared to non-carriers, highlighting a genetic influence on this protein.

“Our research highlights the critical role of TTR levels in predicting heart disease risk,” Arora said. “By understanding the factors that influence TTR levels, we can better identify individuals at high risk and develop targeted interventions to prevent adverse outcomes.”

“These findings underscore the potential benefits of incorporating TTR level measurements in screening programs, especially for individuals with genetic predispositions,” Shetty said.

Arora, the senior author and a cardiologist at the UAB Cardiovascular Institute, says the implications of this study are far-reaching. It suggests that monitoring of TTR levels could be a valuable tool in managing heart disease risk, particularly for those with known genetic variations like the V142I TTR variant. Low TTR levels raise the pre-test probability of a positive genetic test, specifically for detecting the V142I variant, which typically takes time to process.

“This information can be used to counsel family members while they await the results of genetic testing,” Arora said. “This research marks a significant step forward in the quest to understand and mitigate the risks associated with cardiac amyloidosis and other heart-related conditions.”

“In a year that has been the hottest on record to date, these past two weeks have been particularly brutal,” said NASA Administrator Bill Nelson. “Through our over two dozen Earth-observing satellites and over 60 years of data, NASA is providing critical analyses of how our planet is changing and how local communities can prepare, adapt, and stay safe. We are proud to be part of the Biden-Harris Administration efforts to protect communities from extreme heat.”

This preliminary finding comes from data analyses from Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2) and Goddard Earth Observing System Forward Processing (GEOS-FP) systems, which combine millions of global observations from instruments on land, sea, air, and satellites using atmospheric models. GEOS-FP provides rapid, near-real time weather data, while the MERRA-2 climate reanalysis takes longer but ensures the use of best quality observations. These models are run by the Global Modeling and Assimilation Office (GMAO) at NASA’s Goddard Space Flight Center in Greenbelt, Maryland.

Daily global average temperature values from MERRA-2 for the years 1980-2022 are shown in white, values for the year 2023 are shown in pink, and values from 2024 through June are shown in red. Daily global temperature values from July 1 to 23, 2024, from GEOS-FP are shown in purple. The results agree with an independent analysis from the European Union’s Copernicus Earth Observation Programme. While the analyses have small differences, they show broad agreement in the change in temperature over time and hottest days.

The latest daily temperature records follow 13 months of consecutive monthly temperature records, according to scientists from NASA’s Goddard Institute for Space Studies in New York. Their analysis was based on the GISTEMP record, which uses surface instrumental data alone and provides a longer-term view of changes in global temperatures at monthly and annual resolutions going back to the late 19th century.

When deploying a model, users typically strive to ensure its predictions are fair by reducing bias. This often involves techniques like adjusting the features a model uses to make decisions or calibrating the scores it generates.

However, researchers from MIT and Northeastern University argue that these fairness methods are not sufficient to address structural injustices and inherent uncertainties. In a new paper, they show how randomizing a model’s decisions in a structured way can improve fairness in certain situations.

For example, if multiple companies use the same machine-learning model to rank job interview candidates deterministically — without any randomization — then one deserving individual could be the bottom-ranked candidate for every job, perhaps due to how the model weighs answers provided in an online form. Introducing randomization into a model’s decisions could prevent one worthy person or group from always being denied a scarce resource, like a job interview.

Through their analysis, the researchers found that randomization can be especially beneficial when a model’s decisions involve uncertainty or when the same group consistently receives negative decisions.

They present a framework one could use to introduce a specific amount of randomization into a model’s decisions by allocating resources through a weighted lottery. This method, which an individual can tailor to fit their situation, can improve fairness without hurting the efficiency or accuracy of a model.

“Even if you could make fair predictions, should you be deciding these social allocations of scarce resources or opportunities strictly off scores or rankings? As things scale, and we see more and more opportunities being decided by these algorithms, the inherent uncertainties in these scores can be amplified. We show that fairness may require some sort of randomization,” says Shomik Jain, a graduate student in the Institute for Data, Systems, and Society (IDSS) and lead author of the paper.

Jain is joined on the paper by Kathleen Creel, assistant professor of philosophy and computer science at Northeastern University; and senior author Ashia Wilson, the Lister Brothers Career Development Professor in the Department of Electrical Engineering and Computer Science and a principal investigator in the Laboratory for Information and Decision Systems (LIDS). The research will be presented at the International Conference on Machine Learning.

Considering claims

This work builds off a previous paper in which the researchers explored harms that can occur when one uses deterministic systems at scale. They found that using a machine-learning model to deterministically allocate resources can amplify inequalities that exist in training data, which can reinforce bias and systemic inequality.

“Randomization is a very useful concept in statistics, and to our delight, satisfies the fairness demands coming from both a systemic and individual point of view,” Wilson says.

In this paper, they explored the question of when randomization can improve fairness. They framed their analysis around the ideas of philosopher John Broome, who wrote about the value of using lotteries to award scarce resources in a way that honors all claims of individuals.

A person’s claim to a scarce resource, like a kidney transplant, can stem from merit, deservingness, or need. For instance, everyone has a right to life, and their claims on a kidney transplant may stem from that right, Wilson explains.

“When you acknowledge that people have different claims to these scarce resources, fairness is going to require that we respect all claims of individuals. If we always give someone with a stronger claim the resource, is that fair?” Jain says.

That sort of deterministic allocation could cause systemic exclusion or exacerbate patterned inequality, which occurs when receiving one allocation increases an individual’s likelihood of receiving future allocations. In addition, machine-learning models can make mistakes, and a deterministic approach could cause the same mistake to be repeated.

Randomization can overcome these problems, but that doesn’t mean all decisions a model makes should be randomized equally.

Structured randomization

The researchers use a weighted lottery to adjust the level of randomization based on the amount of uncertainty involved in the model’s decision-making. A decision that is less certain should incorporate more randomization.

“In kidney allocation, usually the planning is around projected lifespan, and that is deeply uncertain. If two patients are only five years apart, it becomes a lot harder to measure. We want to leverage that level of uncertainty to tailor the randomization,” Wilson says.

The researchers used statistical uncertainty quantification methods to determine how much randomization is needed in different situations. They show that calibrated randomization can lead to fairer outcomes for individuals without significantly affecting the utility, or effectiveness, of the model.

“There is a balance to be had between overall utility and respecting the rights of the individuals who are receiving a scarce resource, but oftentimes the tradeoff is relatively small,” says Wilson.

However, the researchers emphasize there are situations where randomizing decisions would not improve fairness and could harm individuals, such as in criminal justice contexts.

But there could be other areas where randomization can improve fairness, such as college admissions, and the researchers plan to study other use-cases in future work. They also want to explore how randomization can affect other factors, such as competition or prices, and how it could be used to improve the robustness of machine-learning models.

“We are hoping our paper is a first move toward illustrating that there might be a benefit to randomization. We are offering randomization as a tool. How much you are going to want to do it is going to be up to all the stakeholders in the allocation to decide. And, of course, how they decide is another research question all together,” says Wilson.