Tyler O'Neal, Staff Editor BIOLOGY

UBC's Institute for the Oceans and Fisheries proposes marine heatwaves could wipe out an extra six percent of a country's fish catches, costing millions their jobs

Extremely hot years will wipe out hundreds of thousands of tonnes of fish available for the catch in a country’s waters in this century, on top of projected decreases to fish stocks from long-term climate change, a new UBC study projects. The University of British Columbia is a public research university with campuses in Vancouver and Kelowna, British Columbia, Canada.

Researchers from the UBC Institute for the Oceans and Fisheries (IOF) used a supercomputer model incorporating extreme annual ocean temperatures in Exclusive Economic Zones, where the majority of global fish catches occur, into climate-related projections for fish, fisheries, and their dependent human communities. Marine heatwaves could wipe out an extra six per cent of a country’s fish catches, costing millions their jobs CREDIT Cassiano Psomas on Unsplash

Supercomputer modeling a worst-case scenario where no action is taken to mitigate greenhouse gas emissions they projected a six percent drop in the number of potential catches per year and 77 percent of exploited species are projected to decrease in biomass, or the amount of fish by weight in a given area, due to extremely hot years. These decreases are on top of those projected due to long-term decadal-scale climate change.

The numbers

  • In Pacific Canada, Sockeye salmon catches are projected to decrease by 26 percent on average during a high-temperature event between 2000 and 2050, an annual loss of 260 to 520 tonnes of fish. With losses due to climate change, when a temperature extreme occurs in the 2050s, the total decrease in annual catch would be more than 50 percent or 530 to 1060 tonnes of fish.
  • Peruvian anchoveta catches are projected to decline by 34 percent during an extreme high-temperature event between 2000 and 2050, or more than 900,000 tonnes per year. With climate change, a temperature extreme is projected to cost Peruvian anchoveta fisheries more than 1.5 million tonnes of their potential catch.
  • Overall, a high-temperature extreme event is projected to cause a 25 percent drop in annual revenue for Peruvian anchoveta fisheries, or a loss of around US$600 million
  • Nearly three million jobs in the Indonesian fisheries-related sector are projected to be lost when a high-temperature extreme occurs in their waters between 2000 and 2050.
  • Some stocks are projected to increase due to these extreme events, and climate change, but not enough to mitigate the losses

During extreme ocean temperature events and on top of projected temperature changes each decade, researchers projected that fisheries’ revenues would be cut by an average of three percent globally, and employment by two percent; a potential loss of millions of jobs.

“These extreme annual temperatures will be an additional shock to an overloaded system,” said lead author Dr. William Cheung, professor, and director of UBC’s Institute for the Oceans and Fisheries (IOF). “We see that in the countries where fisheries are already weakened by long-term changes, like ocean warming and deoxygenation, adding the shock of temperature extremes will exacerbate the impacts to a point that will likely exceed the capacity for these fisheries to adapt. It’s not unlike how COVID-19 stresses the healthcare system by adding an extra burden.”

Extreme temperature events are projected to occur more frequently in the future, says co-author Dr. Thomas Frölicher, professor at the climate and environmental physics division of the University of Bern. “Today’s marine heatwaves and their severe impacts on fisheries are bellwethers of the future as these events are generating environmental conditions that long-term global warming will not create for decades.”

Some areas will be worse hit than others, the researchers found, including EEZs in the Indo-Pacific region, particularly waters around South and Southeast Asia, and Pacific Islands; the Eastern Tropical Pacific, and area which runs along the Pacific coast of the Americas; and some countries in the West African region.

In Bangladesh, where fisheries-related sectors employ one-third of the country’s workforce, an extreme marine heat event is expected to cut two percent — about one million — of the country’s fisheries jobs, in addition to the more than six million jobs that will be lost by 2050 due to long-term climate change.

The situation is similarly grim for Ecuador, where extremely high-temperature events are projected to adversely impact an additional 10 percent, or around US$100 million, of the country’s fisheries revenue on top of the 25 percent reduction expected by the mid-21st century.

“This study really highlights the need to develop ways to deal with marine temperature extremes, and soon,” Cheung said. “These temperature extremes are often difficult to predict in terms of when and where they occur, particularly in the hot spots with limited capacity to provide robust scientific predictions for their fisheries. We need to consider that unpredictability when we plan for adaptations to long-term climate change.”

Cheung said that active fisheries management is vital. Potential adaptations include adjusting catch quotas in years when fish stocks are suffering from extreme temperature events, or, in severe cases, shuttering fisheries so that stocks can rebuild. “We need to have mechanisms in place to deal with it,” said Cheung.

It will be important to work with those affected by such adaptation options when developing them, as some decisions could exacerbate impacts on communities’ livelihoods, as well as food and nutrition security, said co-author Dr. Colette Wabnitz, an IOF research associate and lead scientist at the Stanford Center for Ocean Solutions. "Stakeholders are diverse and include not only the industry but also Indigenous communities, small-scale fisheries, and others. They should be involved in discussions about the effects of climate change and marine heatwaves as well as the design and implementation of solutions.”

The study “Marine high-temperature extremes amplify the impacts of climate change on fish and fisheries” was published in the journal Science Advances.

South Korean built deep learning framework enables material design in unseen domain​

Tyler O'Neal, Staff Editor BIOLOGY

KAIST researchers propose a deep neural network-based forward design space exploration using active transfer learning and data augmentation

A new study proposed a deep neural network-based forward design approach that enables an efficient search for superior materials far beyond the domain of the initial training set. This approach compensates for the weak predictive power of neural networks on an unseen domain through gradual updates of the neural network with active transfer learning and data augmentation methods. 

Professor Seungwha Ryu believes that this study will help address a variety of optimization problems that have an astronomical number of possible design configurations. For the grid composite optimization problem, the proposed framework was able to provide excellent designs close to the global optima, even with the addition of a very small dataset corresponding to less than 0.5% of the initial training data-set size. This study was reported in npj Computational Materials last month. KAIST is a national research university located in Daedeok Innopolis, Daejeon, South Korea. 

“We wanted to mitigate the limitation of the neural network, weak predictive power beyond the training set domain for the material or structure design,” said Professor Ryu from the Department of Mechanical Engineering. Figure 1: Schematic of deep learning framework for material design space exploration. Schematic of gradual expansion of reliable prediction domain of DNN based on the addition of data generated from the hyper-heuristic genetic algorithm and active transfer learning.

Neural network-based generative models have been actively investigated as an inverse design method for finding novel materials in a vast design space. However, the applicability of conventional generative models is limited because they cannot access data outside the range of training sets. Advanced generative models that were devised to overcome this limitation also suffer from weak predictive power for the unseen domain.

Professor Ryu’s team, in collaboration with researchers from Professor Grace Gu’s group at UC Berkeley, devised a design method that simultaneously expands the domain using the strong predictive power of a deep neural network and searches for the optimal design by repetitively performing three key steps.

First, it searches for few candidates with improved properties located close to the training set via genetic algorithms, by mixing superior designs within the training set. Then, it checks to see if the candidates have improved properties and expand the training set by duplicating the validated designs via a data augmentation method. Finally, they can expand the reliable prediction domain by updating the neural network with the new superior designs via transfer learning. Because the expansion proceeds along relatively narrow but correct routes toward the optimal design (depicted in the schematic of Fig. 1), the framework enables an efficient search.

As a data-hungry method, a deep neural network model tends to have reliable predictive power only within and near the domain of the training set. When the optimal configuration of materials and structures lies far beyond the initial training set, which frequently is the case, neural network-based design methods suffer from weak predictive power and become inefficient.

Researchers expect that the framework will be applicable for a wide range of optimization problems in other science and engineering disciplines with astronomically large design space because it provides an efficient way of gradually expanding the reliable prediction domain toward the target design while avoiding the risk of being stuck in local minima. Especially, being a less-data-hungry method, design problems in which data generation is time-consuming and expensive will benefit most from this new framework.
The research team is currently applying the optimization framework for the design task of metamaterial structures, segmented thermoelectric generators, and optimal sensor distributions. “From these sets of ongoing studies, we expect to better recognize the pros and cons, and the potential of the suggested algorithm. Ultimately, we want to devise more efficient machine learning-based design approaches,” explained Professor Ryu. This study was funded by the National Research Foundation of Korea and the KAIST Global Singularity Research Project.

Rochester biologists use an evolutionary approach that predicts novel ‘protein partners’ that could contribute to COVID-19 symptoms

Tyler O'Neal, Staff Editor BIOLOGY

COVID-19 not only causes symptoms characteristic of a typical respiratory disorder but has also been known to trigger a wide range of other symptoms in people who had been infected, some lasting even long after individuals test negative for the virus. These symptoms can include abnormal blood clotting, heart damage and failure, kidney disease, brain fog (confusion, memory loss, or difficulty focusing), gastrointestinal problems, and even male infertility.

Yet the mechanisms by which COVID-19 causes these diverse complications remain poorly understood. 

In a new paper published in the journal PeerJ, John (Jack) Werren, the Nathaniel and Helen Wisch Professor of Biology at the University of Rochester, and recent undergraduates Austin Varela ’20 and Sammy Cheng ’21 studied proteins that closely evolve with Angiotensin-converting enzyme 2 (ACE2), the receptor used by the SARS-CoV2 virus to enter human cells.

Using an evolutionary approach, the researchers detected proteins that “coevolve” with ACE2 in mammals as a way to identify networks of proteins that likely interact with ACE2 during its normal functions in the human body. Their rationale is that disruptions caused by COVID-19 in these normal functions of ACE2 could contribute to the unusual pathologies of the disease. Illustration of an Angiotensin-converting enzyme 2 (ACE2) receptor (blue) on a human cell bound to a molecular model of a coronavirus spike (S) protein (red). Using an evolutionary approach, University of Rochester researchers revealed a number of candidate protein partners for ACE2 that could have direct bearing on the acute and chronic complications of COVID-19. (Getty Images)

COVID-19, ACE2, and protein partners

The method used by the researchers revealed a number of candidate protein partners for ACE2 that have not previously been identified as ACE2 interactors, but which could have a direct bearing on the complications experienced by people infected with the virus. These COVID-related complications can include excessive blood clotting as well as an overactive inflammatory response known as a “cytokine storm”—both of which can cause tissue and organ damage.

For example, one hallmark of severe COVID-19 is abnormal blood coagulation throughout the body. The team’s research revealed noteworthy connections between ACE2 and key proteins involved in the coagulation pathway. Another protein, Clusterin, which plays a significant role in “quality control” in the blood by removing misfolded proteins, strongly coevolves with ACE2—implying that they interact with each other biologically. Several proteins involved in cytokine signaling appear to coevolve with ACE2 as well.

“We propose that ACE2 has novel protein interactions that are disrupted during SARS-CoV-2 infection, contributing to the spectrum of COVID-19 pathologies,” Werren says. Finding that ACE2’s evolutionary partners are involved in blood coagulation and cytokine signaling is consistent with this possibility.

“These candidate protein interactions will need to be validated,” Werren says. “But if supported, our findings could inform the development of better treatments and therapeutics for COVID-19 and chronic complications that may arise.”

As an evolutionary geneticist, Werren’s research focuses on the interaction of genomes in symbiotic or parasitic relationships. When the COVID-19 pandemic hit, Werren received a grant from the National Science Foundation’s Rapid Response Research program to study the ACE2’s protein interactions and its network of coevolving proteins. The University’s Nathaniel and Helen Wisch Chair Research Fund, meanwhile, helped support Varela’s and Cheng’s participation. 

“Working on this project was a great opportunity for me,” says Varela, the study’s first author, who started researching protein-protein interactions in the Werren Lab during his sophomore year at Rochester. “Once we uncovered the evolutionary rate partners of ACE2, their potential clinical relevance was immediately clear.”

An evolutionary approach to detecting protein partners and interactions

Werren and his colleagues used a computational and evolutionary approach called evolutionary rate correlation (ERC). The underlying concept is that proteins with similar rates of change during evolution are more likely to have functional interactions. For example, if you look at a phylogenetic tree depicting the evolutionary relationships among mammals, when one protein evolves quickly in a particular species, a protein with which it functionally interacts will tend to evolve quickly as well, and vice versa.

Werren previously applied the ERC method to protein interactions involved in mitochondrial function. Mitochondria are cellular structures that, among other functions, produce energy for the cell. In that study, the ERC method accurately predicted nuclear-encoded proteins known to interact with mitochondria, and also detected proteins not previously known to have a mitochondrial function.

Biomedical researchers currently harness a number of powerful tools for detecting and exploring proteins that interact with each other in biological processes. These include genetic screens, protein coprecipitation, and proteomic profiling.

Evolutionary approaches such as ERC have the potential to become useful additions to the biomedical research toolkit. Werren and his team are now expanding the number of proteins in their analysis. He says, “We hope to produce a more extensive database for researchers so they can explore coevolving partners for proteins of interest in their research.”