Connect with us

The Conversation

Will AI revolutionize drug development? Researchers explain why it depends on how it’s used

Published

on

theconversation.com – Duxin Sun, Associate Dean for Research, Charles Walgreen Jr. Professor of Pharmacy and Pharmaceutical Sciences, University of Michigan – 2025-01-03 07:33:00

A high drug failure rate is more than just a pattern recognition problem.

Thom Leach/Science Photo Library via Getty Images

Duxin Sun, University of Michigan and Christian Macedonia, University of Michigan

The potential of using artificial intelligence in drug discovery and development has sparked both excitement and skepticism among scientists, investors and the general public.

“Artificial intelligence is taking over drug development,” claim some companies and researchers. Over the past few years, interest in using AI to design drugs and optimize clinical trials has driven a surge in research and investment. AI-driven platforms like AlphaFold, which won the 2024 Nobel Prize for its ability to predict the structure of proteins and design new ones, showcase AI’s potential to accelerate drug development.

AI in drug discovery is “nonsense,” warn some industry veterans. They urge that “AI’s potential to accelerate drug discovery needs a reality check,” as AI-generated drugs have yet to demonstrate an ability to address the 90% failure rate of new drugs in clinical trials. Unlike the success of AI in image analysis, its effect on drug development remains unclear.

Pharmacist searching through drawer of drug packages

Behind every drug in your pharmacy are many, many more that failed.

nortonrsx/iStock via Getty Images Plus

We have been following the use of AI in drug development in our work as a pharmaceutical scientist in both academia and the pharmaceutical industry and as a former program manager in the Defense Advanced Research Projects Agency, or DARPA. We argue that AI in drug development is not yet a game-changer, nor is it complete nonsense. AI is not a black box that can turn any idea into gold. Rather, we see it as a tool that, when used wisely and competently, could help address the root causes of drug failure and streamline the process.

Most work using AI in drug development intends to reduce the time and money it takes to bring one drug to market – currently 10 to 15 years and US$1 billion to $2 billion. But can AI truly revolutionize drug development and improve success rates?

AI in drug development

Researchers have applied AI and machine learning to every stage of the drug development process. This includes identifying targets in the body, screening potential candidates, designing drug molecules, predicting toxicity and selecting patients who might respond best to the drugs in clinical trials, among others.

Between 2010 and 2022, 20 AI-focused startups discovered 158 drug candidates, 15 of which advanced to clinical trials. Some of these drug candidates were able to complete preclinical testing in the lab and enter human trials in just 30 months, compared with the typical 3 to 6 years. This accomplishment demonstrates AI’s potential to accelerate drug development.

Drug development is a long and costly process.

On the other hand, while AI platforms may rapidly identify compounds that work on cells in a Petri dish or in animal models, the success of these candidates in clinical trials – where the majority of drug failures occur – remains highly uncertain.

Unlike other fields that have large, high-quality datasets available to train AI models, such as image analysis and language processing, the AI in drug development is constrained by small, low-quality datasets. It is difficult to generate drug-related datasets on cells, animals or humans for millions to billions of compounds. While AlphaFold is a breakthrough in predicting protein structures, how precise it can be for drug design remains uncertain. Minor changes to a drug’s structure can greatly affect its activity in the body and thus how effective it is in treating disease.

Survivorship bias

Like AI, past innovations in drug development like computer-aided drug design, the Human Genome Project and high-throughput screening have improved individual steps of the process in the past 40 years, yet drug failure rates haven’t improved.

Most AI researchers can tackle specific tasks in the drug development process when provided with high-quality data and particular questions to answer. But they are often unfamiliar with the full scope of drug development, reducing challenges into pattern recognition problems and refinement of individual steps of the process. Meanwhile, many scientists with expertise in drug development lack training in AI and machine learning. These communication barriers can hinder scientists from moving beyond the mechanics of current development processes and identifying the root causes of drug failures.

Current approaches to drug development, including those using AI, may have fallen into a survivorship bias trap, overly focusing on less critical aspects of the process while overlooking major problems that contribute most to failure. This is analogous to repairing damage to the wings of aircraft returning from the battle fields in World War II while neglecting the fatal vulnerabilities in engines or cockpits of the planes that never made it back. Researchers often overly focus on how to improve a drug’s individual properties rather than the root causes of failure.

Diagram of airplane with clusters of red dots on the wing tips, tail and cockpit areas

While returning planes might survive hits to the wings, those with damage to the engines or cockpits are less likely to make it back.

Martin Grandjean, McGeddon, US Air Force/Wikimedia Commons, CC BY-SA

The current drug development process operates like an assembly line, relying on a checkbox approach with extensive testing at each step of the process. While AI may be able to reduce the time and cost of the lab-based preclinical stages of this assembly line, it is unlikely to boost success rates in the more costly clinical stages that involve testing in people. The persistent 90% failure rate of drugs in clinical trials, despite 40 years of process improvements, underscores this limitation.

Addressing root causes

Drug failures in clinical trials are not solely due to how these studies are designed; selecting the wrong drug candidates to test in clinical trials is also a major factor. New AI-guided strategies could help address both of these challenges.

Currently, three interdependent factors drive most drug failures: dosage, safety and efficacy. Some drugs fail because they’re too toxic, or unsafe. Other drugs fail because they’re deemed ineffective, often because the dose can’t be increased any further without causing harm.

We and our colleagues propose a machine learning system to help select drug candidates by predicting dosage, safety and efficacy based on five previously overlooked features of drugs. Specifically, researchers could use AI models to determine how specifically and potently the drug binds to known and unknown targets, the level of these targets in the body, how concentrated the drug becomes in healthy and diseased tissues, and the drug’s structural properties.

These features of AI-generated drugs could be tested in what we call phase 0+ trials, using ultra-low doses in patients with severe and mild disease. This could help researchers identify optimal drugs while reducing the costs of the current “test-and-see” approach to clinical trials.

While AI alone might not revolutionize drug development, it can help address the root causes of why drugs fail and streamline the lengthy process to approval.The Conversation

Duxin Sun, Associate Dean for Research, Charles Walgreen Jr. Professor of Pharmacy and Pharmaceutical Sciences, University of Michigan and Christian Macedonia, Adjunct Professor in Pharmaceutical Sciences, University of Michigan

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Read More

The post Will AI revolutionize drug development? Researchers explain why it depends on how it’s used appeared first on theconversation.com

The Conversation

Microbes can colonize space, produce drugs and create energy − researchers are simulating their inner workings to harness how

Published

on

theconversation.com – Blaise Manga Enuh, Postdoctoral Research Associate in Microbial Genomics and Systems Biology, University of Wisconsin-Madison – 2025-01-06 07:19:00

Genome-scale metabolic models capture the complex chemical reactions that allow cells to function.
Yuri Arcurs/iStock via Getty Images Plus

Blaise Manga Enuh, University of Wisconsin-Madison

After so many years learning how microbes work, researchers are now digitally recreating their inner workings to tackle challenges ranging from climate change to space colonization.

In my work as a computational biologist, I research ways to get microbes to produce more useful chemicals, such as fuels and bioplastics, that can be used in the energy, agricultural or pharmaceutical industries. Traditionally, researchers have to conduct several trial-and-error experiments on petri dishes in order to determine the optimal conditions microbes need to produce high amounts of chemicals.

Instead, I am able to simulate these experiments all from behind a computer screen through digital blueprints that replicate the inside of microbes. Called genome-scale metabolic models, or GEMs, these virtual labs significantly reduce the time and cost required to figure out what researchers need to do to get what they’re looking for. With GEMs, researchers cannot only explore the complex network of metabolic pathways that allow living organisms to function, but also tweak, test and predict how microbes would behave in different environments, including on other planets.

As GEM technology continues to evolve, I believe these models will play an increasingly important role in shaping the future of biotechnology, medicine and space exploration.

What are genome-scale metabolic models?

Genome-scale metabolic models are digital maps of all the known chemical reactions that occur in cells – that is, the cell’s metabolism. These reactions are crucial for converting food into energy, building cellular structures and detoxifying harmful substances.

To create a GEM, I begin by analyzing an organism’s genome, which contains the genetic instructions cells use to produce proteins. A type of protein coded in the genome called enzymes are the workhorses of metabolism – they facilitate the conversion of nutrients into energy and building blocks for cells.

By linking the genes that encode enzymes to the chemical reactions they help make happen, I can build a comprehensive model that maps out the connections between genes, reactions and metabolites.

Once I build a GEM, I use some advanced computational simulations to make it work like a live cell or microbe would. One of the most common algorithms researchers use to do these simulations is called a flux balance analysis. This mathematical algorithm analyzes available data about metabolism, then makes predictions on how different chemical reactions and metabolites would act under specific conditions.

This makes GEMs particularly useful for understanding how organisms respond to genetic changes and environmental stresses. For example, I can use this method to predict how an organism will react when a specific gene is knocked out. I could also use it to predict how it might adapt to the presence of different chemicals in its environment or a lack of food.

Solving energy and climate challenges

Most of the chemicals used in agriculture, pharmaceuticals and fuels are obtained from fossil fuels. However, fossil fuels are a limited resource and significantly contribute to climate change.

Instead of extracting energy from fossil fuels, my team at the Great Lakes Bioenergy Research Center of the University of Wisconsin-Madison focuses on developing sustainable biofuels and bioproducts from plant waste. This includes cornstalk after the ears are harvested, nonedible plants such as grass, and algae. We study which crop wastes can be used for bioenergy, how to use microbes to convert them into energy, and ways to sustainably manage the land on which those crops are grown.

I am building a genome-scale metabolic model for Novosphingobium aromaticivorans, a species of bacteria that can convert very complex chemicals in plant waste to chemicals that are valuable to people, such as those used to make bioplastics, pharmaceuticals and fuels. With a clearer understanding of this conversion process, I can improve the model to more accurately simulate the conditions needed to synthesize greater amounts of these chemicals.

Researchers can then replicate these conditions in real life to generate materials that are cheaper and more accessible than those made from fossil fuels.

Bioinformatics analyzes biological data to answer questions about living organisms.

Extreme microbes and space colonization

There are microbes on Earth that can survive in extremely harsh environments. For example, Chromohalobacter canadensis can live in extremely salty conditions. Similarly, Alicyclobacillus tolerans can thrive in very acidic environments.

Since other planets typically have similarly harsh climates, these microbes may not only be able to thrive and reproduce on these planets but could potentially change the environment so humans can live there as well.

Combining GEMs with machine learning, I saw that C. canadensis and A. tolerans can undergo chemical changes that help them survive in extreme conditions. They have special proteins in their cell walls that work with enzymes to balance the chemicals in their internal environment with the chemicals in their external environment.

With GEMs, scientists can simulate the environments of other planets to study how microbes survive without necessarily needing to go to those planets themselves.

The future of GEMs

Every day, researchers are generating large amounts of data about microbial metabolism. As GEM technology advances, it opens the door to exciting new possibilities in medicine, energy, space and other areas.

Synthetic biologists can use GEMs to design entirely new organisms or metabolic pathways from scratch. This field could advance biomanufacturing by enabling the creation of organisms that efficiently produce new materials, drugs or even food.

Whole human body GEMs can also serve as an atlas for the metabolics of complex diseases. They can help map how the chemical environment of the body changes with obesity or diabetes.

Whether it’s producing biofuels or engineering new organisms, GEMs provide a powerful tool for both basic research and industrial applications. As computational biology and GEMs advance, these technologies will continue to transform how scientists understand and manipulate the metabolisms of living organisms.The Conversation

Blaise Manga Enuh, Postdoctoral Research Associate in Microbial Genomics and Systems Biology, University of Wisconsin-Madison

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Read More

The post Microbes can colonize space, produce drugs and create energy − researchers are simulating their inner workings to harness how appeared first on theconversation.com

Continue Reading

The Conversation

Americans’ rage at insurers goes beyond health coverage – the author of ‘Delay, Deny, Defend’ points to 3 reforms that could help

Published

on

theconversation.com – Jay Feinman, Distinguished Professor of Law Emeritus, Rutgers University – 2025-01-06 11:06:00

Jay Feinman, Rutgers University

My book “Delay, Deny, Defend: Why Insurance Companies Don’t Pay Claims and What You Can Do About It” was thrust into the spotlight recently, after UnitedHealthcare CEO Brian Thompson was shot and killed in what authorities say was a targeted attack outside the company’s annual investors conference. Investigators at the scene found bullet casings inscribed with the words “delay,” “deny” and “depose.”

The unsettling echo of the book’s title struck me and many others.

That killing – and the torrent of online outrage that followed – put Americans’ unhappiness with health insurers at the front of the national conversation. Many people responded not by mourning Thompson, but by blaming UnitedHealthcare and other insurers for failing to pay for essential medical treatments. Gleeful online trolls even celebrated the alleged killer as a heroic vigilante.

Speaking as an insurance scholar, I think few should be surprised by this ghoulish reaction. The killing revealed many Americans’ resentment and even rage about insurance companies. And while the focus has been on health insurance, these frustrations extend across the broader insurance landscape. Homeowners insurance, for example, is becoming harder to get in many states even as coverage is shrinking, and auto insurance rates are skyrocketing. These trends are fueling widespread discontent with insurers of all kinds.

Why policyholders feel betrayed

As many recent stories of health insurance denials in the news show, policyholders are most outraged when insurers fail to keep their promises to pay claims promptly and fairly.

And as I read people’s stories about their own experiences, I kept hearing echoes from my book. Too often, people say, insurance companies delay paying some claims, deny other valid claims altogether, and force policyholders to defend themselves in court – all to increase profits by cutting claim costs.

But problems often begin long before anyone files a claim. Insurance consumers generally don’t know much about what they are buying. For homeowners, auto and many other types of insurance, companies seldom provide copies of policy language or accessible summaries of policy terms to prospective policyholders.

A vintage illustration of a confident-looking businessman saying taken from an insurance advertisement circa 1942.

Insurance advertisements, like this one from the early 1940s, often sell the promise of safety and security.

GraphicaArtis/Hulton Archive via Getty Images

Even when consumers have access to policies, many don’t read or can’t understand the long, complex legal documents. Similarly, they can’t anticipate the many ways a loss could occur or the problems that could result if it does. As a result, they are only aware of a few key terms and otherwise believe that they will be “in good hands” with a “good neighbor,” to quote two of the iconic phrases of insurance advertising.

Then, when consumers need coverage, they discover that there are significant protection gaps. Health insurance can involve a tangle of limitations due to provider networks, medical necessity rules and preauthorization requirements. Homeowners reasonably expect that they will be fully covered for all major losses, but insurers have cut back coverage to account for rising costs due to inflation and climate change.

As a result, when disaster strikes, too many Americans feel like they haven’t gotten the security they already paid for.

An insurance industry Americans can trust

Rebuilding trust in insurance won’t be easy, but it’s essential. Insurance is the great protector of financial security for the American middle class, but only when it works. As the recent reaction demonstrates, it needs to work better. The insurance industry won’t change by itself; the financial pressures on insurers from increasing losses and fierce market competition are too great.

In order for insurance to serve its goals, lawmakers and regulators will need to take action. Based on my research, I see three big areas for improvement.

Two young men hold handmade signs that read

Protestors hold signs outside Manhattan Criminal Court in New York on Dec. 23, 2024, after suspect Luigi Mangione appeared for his arraignment on state murder charges in the killing of UnitedHealthcare CEO Brian Thompson.

Andrew Lichtenstein/Corbis via Getty Images

First, the government can help make the market for insurance work better. Markets need information, and better information produces better results. Regulators should require that key information about coverage be available in an accessible format for all types of insurance.

Consumers also need information on the quality of companies offering policies, and whether a company pays claims promptly and fairly is a key measure of quality. Consumers don’t have access to much reliable information on that now, so disclosure should be mandated there as well.

Second, states would be wise to consider minimum coverage standards, especially for homeowners insurance, as insurers have been cutting back on coverage recently to reduce costs. New York addressed a similar problem in 1943, legislatively adopting a Standard Fire Policy, since copied in many states.

Some 70 years later, the Affordable Care Act did something similar by requiring that insurers cover 10 “Essential Health Benefits.” In both cases, lawmakers set minimum standards that every company must meet. States again need to consider whether insurance coverage is too important to be left purely to the vagaries of the market.

Third, policyholders need effective remedies when insurance companies are found to have acted unreasonably. Many insurance claims result in good-faith disputes about how much the insurance company should pay — for example, whether roof damage was caused by hail, which is usually covered by insurance, or just wear and tear, which isn’t. But other times, insurance companies deny claims after inadequate investigations or for spurious reasons.

For example, a 2023 Washington Post investigation concluded that in the wake of Hurricane Ian, some Florida insurance companies aggressively sought to limit payouts by altering the work of their adjusters who inspected damaged homes. Some policyholders and their families had their Hurricane Ian claims reduced by 45% to 97%. The American Policyholder Association, a nonprofit insurance industry watchdog group, claimed to find “compelling evidence of what appears to be multiple instances of systematic criminal fraud perpetrated to cheat policyholders out of fair insurance claims.”

When people find themselves in this sort of situation, they have to spend lots of time and effort fighting to get what they were owed in the first place. Even when an insurance company eventually relents, it still hasn’t fulfilled its original promise to the policyholder to settle claims promptly and fairly. In these cases, requiring additional compensation to policyholders and insurer disincentives for unreasonable conduct would level the playing field.

The deep resentment many Americans feel toward insurance companies became apparent after the killing of Brian Thompson. Reforms such as these would be a meaningful response to that resentment.The Conversation

Jay Feinman, Distinguished Professor of Law Emeritus, Rutgers University

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Read More

The post Americans’ rage at insurers goes beyond health coverage – the author of ‘Delay, Deny, Defend’ points to 3 reforms that could help appeared first on theconversation.com

Continue Reading

The Conversation

How Christian nationalism played a role in incorporating the phrase ‘so help me God’ in the presidential oath of office

Published

on

theconversation.com – David B. Parker, Professor of History, Kennesaw State University – 2025-01-06 07:23:00

An oil painting of George Washington taking the oath of office as the first president of the United States on April 30, 1789, in New York City.

Ramon de Elorriaga/Encyclopedia Britannica via Wikimedia Commons

David B. Parker, Kennesaw State University

On Jan. 20, 2025, Donald Trump will take the presidential oath of office: “I do solemnly swear that I will faithfully execute the Office of President of the United States, and will to the best of my ability, preserve, protect and defend the Constitution of the United States.” And then he will probably add the phrase “so help me God.”

Those four little words are not in the Constitution,

but for many Americans, the phrase has been a part of the oath ever since George Washington was said to have added it 236 years ago.

But did Washington really say “so help me God?” There is no evidence that he did. In fact, no one said he did until 1854, 65 years later, when Rufus Griswold, an editor and literary critic, told the story in a book titled “The Republican Court”: “[Washington] added, with fervor, his eyes closed, that his whole soul might be absorbed in the supplication, ‘So help me God!’”

As a professor of U.S. history, I don’t care if Washington said it or not; my interest is in how quickly “so help me God” became established in the American national memory.

For a 2014 article titled “In Griswold We Trust,” I used various online databases such as Google Books, Internet Archive, American Periodicals Series and Newspapers.com to search for the phrase. Before 1854, there are no accounts of Washington saying “so help me God” at the end of the oath – at least in the millions of print records covered by the databases. Then Griswold told the story, and by the end of the 1850s, almost a dozen books and magazine articles had repeated it. Griswold’s story was so thoroughly accepted that, through the 20th century, no one, including academic scholars, thought to question it.

The best way to understand Griswold’s mythic insertion of “so help me God” into the presidential oath is through the lens of Christian nationalism. While the phrase is relatively new, Christian nationalism itself has been around for a long time.

Second Great Awakening

Scholars Andrew Whitehead and Samuel Perry have defined Christian nationalism as “a cultural framework … that idealizes and advocates a fusion of Christianity with American civic life.”

Christian nationalism was big in the early 19th century. Legal scholar Steven K. Green noted in his 2015 book, “Inventing a Christian America,” that the Second Great Awakening, a Protestant evangelical revival movement that peaked in the 1830s, “brought about … a desire to see religious values reflected in the nation’s culture and institutions.”

Rev. Ezra Stiles Ely, a Presbyterian minister in Philadelphia, took things a step further when he told his congregation in 1828 that only leaders “known to be avowedly Christians” should be elected.

In the words of religious studies scholar Richard Hughes, many participants in the Second Great Awakening “sought to transform the nation into a Christian Republic.” In its aftermath, Griswold’s account of Washington prayerfully adding “so help me God” to the presidential oath became part of America’s Christian creation myth.

Another age, another “so help me God” story

Like many cultural ideas, Christian nationalism has waxed and waned through American history. It was popular again in the years just after World War II, a time of increased tensions between the United States and the “godless communists” of the Soviet Union.

Religion was an important weapon in the Cold War. As Sen. Joseph McCarthy said, “The fate of the world rests with the clash between the atheism of Moscow and the Christian spirit throughout other parts of the world.”

In this Cold War context, the U.S. added “under God” to the Pledge of Allegiance, made “In God We Trust” the country’s national motto and created a new version of the Griswold story: that every president, not just Washington, had ended their oath of office with “so help me God.”

A poster with the Pledge of Allegiance next to the American flag.

The Pledge of Allegiance.

United States Government Publishing Office via Wikimedia Commons

Actually, there is no compelling evidence that any president added “so help me God” before September 1881, when Chester A. Arthur was sworn in after the death of James Garfield.

But it was important in Cold War America to prove that it was a Christian nation, so a new story was added to the American creation myth: Through the nation’s history, all presidents invoked God as part of their oath.

A search of the databases shows that this story began in 1948. One of the earliest examples was from Frank Waldrop, editor of the Washington Times-Herald, responding to the Supreme Court’s decision in McCollum v. Board of Education that it was unconstitutional for public schools to promote religion. “Every President from Washington down to Harry Truman has always taken that oath with his hand on the Bible,” Waldrop wrote, “and every President … has added the undeniably religious phrase, ‘So help me, God.’”

Waldrop used the assertion that presidents have all said “so help me God” as an argument for inserting religion into public schools. This is an important point about Christian nationalism: As scholar Eric McDaniel and others have shown, it is not just a view of the past; it is a call for action, specifically to reclaim America as a “holy land.”

Christian nationalism relies on a flawed understanding of the American past, but it has become an increasingly important part of our history.The Conversation

David B. Parker, Professor of History, Kennesaw State University

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Read More

The post How Christian nationalism played a role in incorporating the phrase ‘so help me God’ in the presidential oath of office appeared first on theconversation.com

Continue Reading

Trending