Genetically Sequencing Healthy Babies Yielded Surprising Results
Today in Melrose, Massachusetts, Cora Stetson is the picture of good health, a bubbly precocious 2-year-old. But Cora has two separate mutations in the gene that produces a critical enzyme called biotinidase and her body produces only 40 percent of the normal levels of that enzyme.
In the last few years, the dream of predicting and preventing diseases through genomics, starting in childhood, is finally within reach.
That's enough to pass conventional newborn (heelstick) screening, but may not be enough for normal brain development, putting baby Cora at risk for seizures and cognitive impairment. But thanks to an experimental study in which Cora's DNA was sequenced after birth, this condition was discovered and she is being treated with a safe and inexpensive vitamin supplement.
Stories like these are beginning to emerge from the BabySeq Project, the first clinical trial in the world to systematically sequence healthy newborn infants. This trial was led by my research group with funding from the National Institutes of Health. While still controversial, it is pointing the way to a future in which adults, or even newborns, can receive comprehensive genetic analysis in order to determine their risk of future disease and enable opportunities to prevent them.
Some believe that medicine is still not ready for genomic population screening, but others feel it is long overdue. After all, the sequencing of the Human Genome Project was completed in 2003, and with this milestone, it became feasible to sequence and interpret the genome of any human being. The costs have come down dramatically since then; an entire human genome can now be sequenced for about $800, although the costs of bioinformatic and medical interpretation can add another $200 to $2000 more, depending upon the number of genes interrogated and the sophistication of the interpretive effort.
Two-year-old Cora Stetson, whose DNA sequencing after birth identified a potentially dangerous genetic mutation in time for her to receive preventive treatment.
(Photo courtesy of Robert Green)
The ability to sequence the human genome yielded extraordinary benefits in scientific discovery, disease diagnosis, and targeted cancer treatment. But the ability of genomes to detect health risks in advance, to actually predict the medical future of an individual, has been mired in controversy and slow to manifest. In particular, the oft-cited vision that healthy infants could be genetically tested at birth in order to predict and prevent the diseases they would encounter, has proven to be far tougher to implement than anyone anticipated.
But in the last few years, the dream of predicting and preventing diseases through genomics, starting in childhood, is finally within reach. Why did it take so long? And what remains to be done?
Great Expectations
Part of the problem was the unrealistic expectations that had been building for years in advance of the genomic science itself. For example, the 1997 film Gattaca portrayed a near future in which the lifetime risk of disease was readily predicted the moment an infant is born. In the fanfare that accompanied the completion of the Human Genome Project, the notion of predicting and preventing future disease in an individual became a powerful meme that was used to inspire investment and public support for genomic research long before the tools were in place to make it happen.
Another part of the problem was the success of state-mandated newborn screening programs that began in the 1960's with biochemical tests of the "heel-stick" for babies with metabolic disorders. These programs have worked beautifully, costing only a few dollars per baby and saving thousands of infants from death and severe cognitive impairment. It seemed only logical that a new technology like genome sequencing would add power and promise to such programs. But instead of embracing the notion of newborn sequencing, newborn screening laboratories have thus far rejected the entire idea as too expensive, too ambiguous, and too threatening to the comfortable constituency that they had built within the public health framework.
"What can you find when you look as deeply as possible into the medical genomes of healthy individuals?"
Creating the Evidence Base for Preventive Genomics
Despite a number of obstacles, there are researchers who are exploring how to achieve the original vision of genomic testing as a tool for disease prediction and prevention. For example, in our NIH-funded MedSeq Project, we were the first to ask the question: "What can you find when you look as deeply as possible into the medical genomes of healthy individuals?"
Most people do not understand that genetic information comes in four separate categories: 1) dominant mutations putting the individual at risk for rare conditions like familial forms of heart disease or cancer, (2) recessive mutations putting the individual's children at risk for rare conditions like cystic fibrosis or PKU, (3) variants across the genome that can be tallied to construct polygenic risk scores for common conditions like heart disease or type 2 diabetes, and (4) variants that can influence drug metabolism or predict drug side effects such as the muscle pain that occasionally occurs with statin use.
The technological and analytical challenges of our study were formidable, because we decided to systematically interrogate over 5000 disease-associated genes and report results in all four categories of genetic information directly to the primary care physicians for each of our volunteers. We enrolled 200 adults and found that everyone who was sequenced had medically relevant polygenic and pharmacogenomic results, over 90 percent carried recessive mutations that could have been important to reproduction, and an extraordinary 14.5 percent carried dominant mutations for rare genetic conditions.
A few years later we launched the BabySeq Project. In this study, we restricted the number of genes to include only those with child/adolescent onset that could benefit medically from early warning, and even so, we found 9.4 percent carried dominant mutations for rare conditions.
At first, our interpretation around the high proportion of apparently healthy individuals with dominant mutations for rare genetic conditions was simple – that these conditions had lower "penetrance" than anticipated; in other words, only a small proportion of those who carried the dominant mutation would get the disease. If this interpretation were to hold, then genetic risk information might be far less useful than we had hoped.
Suddenly the information available in the genome of even an apparently healthy individual is looking more robust, and the prospect of preventive genomics is looking feasible.
But then we circled back with each adult or infant in order to examine and test them for any possible features of the rare disease in question. When we did this, we were surprised to see that in over a quarter of those carrying such mutations, there were already subtle signs of the disease in question that had not even been suspected! Now our interpretation was different. We now believe that genetic risk may be responsible for subclinical disease in a much higher proportion of people than has ever been suspected!
Meanwhile, colleagues of ours have been demonstrating that detailed analysis of polygenic risk scores can identify individuals at high risk for common conditions like heart disease. So adding up the medically relevant results in any given genome, we start to see that you can learn your risks for a rare monogenic condition, a common polygenic condition, a bad effect from a drug you might take in the future, or for having a child with a devastating recessive condition. Suddenly the information available in the genome of even an apparently healthy individual is looking more robust, and the prospect of preventive genomics is looking feasible.
Preventive Genomics Arrives in Clinical Medicine
There is still considerable evidence to gather before we can recommend genomic screening for the entire population. For example, it is important to make sure that families who learn about such risks do not suffer harms or waste resources from excessive medical attention. And many doctors don't yet have guidance on how to use such information with their patients. But our research is convincing many people that preventive genomics is coming and that it will save lives.
In fact, we recently launched a Preventive Genomics Clinic at Brigham and Women's Hospital where information-seeking adults can obtain predictive genomic testing with the highest quality interpretation and medical context, and be coached over time in light of their disease risks toward a healthier outcome. Insurance doesn't yet cover such testing, so patients must pay out of pocket for now, but they can choose from a menu of genetic screening tests, all of which are more comprehensive than consumer-facing products. Genetic counseling is available but optional. So far, this service is for adults only, but sequencing for children will surely follow soon.
As the costs of sequencing and other Omics technologies continue to decline, we will see both responsible and irresponsible marketing of genetic testing, and we will need to guard against unscientific claims. But at the same time, we must be far more imaginative and fast moving in mainstream medicine than we have been to date in order to claim the emerging benefits of preventive genomics where it is now clear that suffering can be averted, and lives can be saved. The future has arrived if we are bold enough to grasp it.
Funding and Disclosures:
Dr. Green's research is supported by the National Institutes of Health, the Department of Defense and through donations to The Franca Sozzani Fund for Preventive Genomics. Dr. Green receives compensation for advising the following companies: AIA, Applied Therapeutics, Helix, Ohana, OptraHealth, Prudential, Verily and Veritas; and is co-founder and advisor to Genome Medical, Inc, a technology and services company providing genetics expertise to patients, providers, employers and care systems.
Researchers probe extreme gene therapy for severe alcoholism
Story by Freethink
A single shot — a gene therapy injected into the brain — dramatically reduced alcohol consumption in monkeys that previously drank heavily. If the therapy is safe and effective in people, it might one day be a permanent treatment for alcoholism for people with no other options.
The challenge: Alcohol use disorder (AUD) means a person has trouble controlling their alcohol consumption, even when it is negatively affecting their life, job, or health.
In the U.S., more than 10 percent of people over the age of 12 are estimated to have AUD, and while medications, counseling, or sheer willpower can help some stop drinking, staying sober can be a huge struggle — an estimated 40-60 percent of people relapse at least once.
A team of U.S. researchers suspected that an in-development gene therapy for Parkinson’s disease might work as a dopamine-replenishing treatment for alcoholism, too.
According to the CDC, more than 140,000 Americans are dying each year from alcohol-related causes, and the rate of deaths has been rising for years, especially during the pandemic.
The idea: For occasional drinkers, alcohol causes the brain to release more dopamine, a chemical that makes you feel good. Chronic alcohol use, however, causes the brain to produce, and process, less dopamine, and this persistent dopamine deficit has been linked to alcohol relapse.
There is currently no way to reverse the changes in the brain brought about by AUD, but a team of U.S. researchers suspected that an in-development gene therapy for Parkinson’s disease might work as a dopamine-replenishing treatment for alcoholism, too.
To find out, they tested it in heavy-drinking monkeys — and the animals’ alcohol consumption dropped by 90% over the course of a year.
How it works: The treatment centers on the protein GDNF (“glial cell line-derived neurotrophic factor”), which supports the survival of certain neurons, including ones linked to dopamine.
For the new study, a harmless virus was used to deliver the gene that codes for GDNF into the brains of four monkeys that, when they had the option, drank heavily — the amount of ethanol-infused water they consumed would be equivalent to a person having nine drinks per day.
“We targeted the cell bodies that produce dopamine with this gene to increase dopamine synthesis, thereby replenishing or restoring what chronic drinking has taken away,” said co-lead researcher Kathleen Grant.
To serve as controls, another four heavy-drinking monkeys underwent the same procedure, but with a saline solution delivered instead of the gene therapy.
The results: All of the monkeys had their access to alcohol removed for two months following the surgery. When it was then reintroduced for four weeks, the heavy drinkers consumed 50 percent less compared to the control group.
When the researchers examined the monkeys’ brains at the end of the study, they were able to confirm that dopamine levels had been replenished in the treated animals, but remained low in the controls.
The researchers then took the alcohol away for another four weeks, before giving it back for four. They repeated this cycle for a year, and by the end of it, the treated monkeys’ consumption had fallen by more than 90 percent compared to the controls.
“Drinking went down to almost zero,” said Grant. “For months on end, these animals would choose to drink water and just avoid drinking alcohol altogether. They decreased their drinking to the point that it was so low we didn’t record a blood-alcohol level.”
When the researchers examined the monkeys’ brains at the end of the study, they were able to confirm that dopamine levels had been replenished in the treated animals, but remained low in the controls.
Looking ahead: Dopamine is involved in a lot more than addiction, so more research is needed to not only see if the results translate to people but whether the gene therapy leads to any unwanted changes to mood or behavior.
Because the therapy requires invasive brain surgery and is likely irreversible, it’s unlikely to ever become a common treatment for alcoholism — but it could one day be the only thing standing between people with severe AUD and death.
“[The treatment] would be most appropriate for people who have already shown that all our normal therapeutic approaches do not work for them,” said Grant. “They are likely to create severe harm or kill themselves or others due to their drinking.”
This article originally appeared on Freethink, home of the brightest minds and biggest ideas of all time.
Massive benefits of AI come with environmental and human costs. Can AI itself be part of the solution?
The recent explosion of generative artificial intelligence tools like ChatGPT and Dall-E enabled anyone with internet access to harness AI’s power for enhanced productivity, creativity, and problem-solving. With their ever-improving capabilities and expanding user base, these tools proved useful across disciplines, from the creative to the scientific.
But beneath the technological wonders of human-like conversation and creative expression lies a dirty secret—an alarming environmental and human cost. AI has an immense carbon footprint. Systems like ChatGPT take months to train in high-powered data centers, which demand huge amounts of electricity, much of which is still generated with fossil fuels, as well as water for cooling. “One of the reasons why Open AI needs investments [to the tune of] $10 billion from Microsoft is because they need to pay for all of that computation,” says Kentaro Toyama, a computer scientist at the University of Michigan. There’s also an ecological toll from mining rare minerals required for hardware and infrastructure. This environmental exploitation pollutes land, triggers natural disasters and causes large-scale human displacement. Finally, for data labeling needed to train and correct AI algorithms, the Big Data industry employs cheap and exploitative labor, often from the Global South.
Generative AI tools are based on large language models (LLMs), with most well-known being various versions of GPT. LLMs can perform natural language processing, including translating, summarizing and answering questions. They use artificial neural networks, called deep learning or machine learning. Inspired by the human brain, neural networks are made of millions of artificial neurons. “The basic principles of neural networks were known even in the 1950s and 1960s,” Toyama says, “but it’s only now, with the tremendous amount of compute power that we have, as well as huge amounts of data, that it’s become possible to train generative AI models.”
Though there aren’t any official figures about the power consumption or emissions from data centers, experts estimate that they use one percent of global electricity—more than entire countries.
In recent months, much attention has gone to the transformative benefits of these technologies. But it’s important to consider that these remarkable advances may come at a price.
AI’s carbon footprint
In their latest annual report, 2023 Landscape: Confronting Tech Power, the AI Now Institute, an independent policy research entity focusing on the concentration of power in the tech industry, says: “The constant push for scale in artificial intelligence has led Big Tech firms to develop hugely energy-intensive computational models that optimize for ‘accuracy’—through increasingly large datasets and computationally intensive model training—over more efficient and sustainable alternatives.”
Though there aren’t any official figures about the power consumption or emissions from data centers, experts estimate that they use one percent of global electricity—more than entire countries. In 2019, Emma Strubell, then a graduate researcher at the University of Massachusetts Amherst, estimated that training a single LLM resulted in over 280,000 kg in CO2 emissions—an equivalent of driving almost 1.2 million km in a gas-powered car. A couple of years later, David Patterson, a computer scientist from the University of California Berkeley, and colleagues, estimated GPT-3’s carbon footprint at over 550,000 kg of CO2 In 2022, the tech company Hugging Face, estimated the carbon footprint of its own language model, BLOOM, as 25,000 kg in CO2 emissions. (BLOOM’s footprint is lower because Hugging Face uses renewable energy, but it doubled when other life-cycle processes like hardware manufacturing and use were added.)
Luckily, despite the growing size and numbers of data centers, their increasing energy demands and emissions have not kept pace proportionately—thanks to renewable energy sources and energy-efficient hardware.
But emissions don’t tell the full story.
AI’s hidden human cost
“If historical colonialism annexed territories, their resources, and the bodies that worked on them, data colonialism’s power grab is both simpler and deeper: the capture and control of human life itself through appropriating the data that can be extracted from it for profit.” So write Nick Couldry and Ulises Mejias, authors of the book The Costs of Connection.
The energy requirements, hardware manufacture and the cheap human labor behind AI systems disproportionately affect marginalized communities.
Technologies we use daily inexorably gather our data. “Human experience, potentially every layer and aspect of it, is becoming the target of profitable extraction,” Couldry and Meijas say. This feeds data capitalism, the economic model built on the extraction and commodification of data. While we are being dispossessed of our data, Big Tech commodifies it for their own benefit. This results in consolidation of power structures that reinforce existing race, gender, class and other inequalities.
“The political economy around tech and tech companies, and the development in advances in AI contribute to massive displacement and pollution, and significantly changes the built environment,” says technologist and activist Yeshi Milner, who founded Data For Black Lives (D4BL) to create measurable change in Black people’s lives using data. The energy requirements, hardware manufacture and the cheap human labor behind AI systems disproportionately affect marginalized communities.
AI’s recent explosive growth spiked the demand for manual, behind-the-scenes tasks, creating an industry described by Mary Gray and Siddharth Suri as “ghost work” in their book. This invisible human workforce that lies behind the “magic” of AI, is overworked and underpaid, and very often based in the Global South. For example, workers in Kenya who made less than $2 an hour, were the behind the mechanism that trained ChatGPT to properly talk about violence, hate speech and sexual abuse. And, according to an article in Analytics India Magazine, in some cases these workers may not have been paid at all, a case for wage theft. An exposé by the Washington Post describes “digital sweatshops” in the Philippines, where thousands of workers experience low wages, delays in payment, and wage theft by Remotasks, a platform owned by Scale AI, a $7 billion dollar American startup. Rights groups and labor researchers have flagged Scale AI as one company that flouts basic labor standards for workers abroad.
It is possible to draw a parallel with chattel slavery—the most significant economic event that continues to shape the modern world—to see the business structures that allow for the massive exploitation of people, Milner says. Back then, people got chocolate, sugar, cotton; today, they get generative AI tools. “What’s invisible through distance—because [tech companies] also control what we see—is the massive exploitation,” Milner says.
“At Data for Black Lives, we are less concerned with whether AI will become human…[W]e’re more concerned with the growing power of AI to decide who’s human and who’s not,” Milner says. As a decision-making force, AI becomes a “justifying factor for policies, practices, rules that not just reinforce, but are currently turning the clock back generations years on people’s civil and human rights.”
Ironically, AI plays an important role in mitigating its own harms—by plowing through mountains of data about weather changes, extreme weather events and human displacement.
Nuria Oliver, a computer scientist, and co-founder and vice-president of the European Laboratory of Learning and Intelligent Systems (ELLIS), says that instead of focusing on the hypothetical existential risks of today’s AI, we should talk about its real, tangible risks.
“Because AI is a transverse discipline that you can apply to any field [from education, journalism, medicine, to transportation and energy], it has a transformative power…and an exponential impact,” she says.
AI's accountability
“At the core of what we were arguing about data capitalism [is] a call to action to abolish Big Data,” says Milner. “Not to abolish data itself, but the power structures that concentrate [its] power in the hands of very few actors.”
A comprehensive AI Act currently negotiated in the European Parliament aims to rein Big Tech in. It plans to introduce a rating of AI tools based on the harms caused to humans, while being as technology-neutral as possible. That sets standards for safe, transparent, traceable, non-discriminatory, and environmentally friendly AI systems, overseen by people, not automation. The regulations also ask for transparency in the content used to train generative AIs, particularly with copyrighted data, and also disclosing that the content is AI-generated. “This European regulation is setting the example for other regions and countries in the world,” Oliver says. But, she adds, such transparencies are hard to achieve.
Google, for example, recently updated its privacy policy to say that anything on the public internet will be used as training data. “Obviously, technology companies have to respond to their economic interests, so their decisions are not necessarily going to be the best for society and for the environment,” Oliver says. “And that’s why we need strong research institutions and civil society institutions to push for actions.” ELLIS also advocates for data centers to be built in locations where the energy can be produced sustainably.
Ironically, AI plays an important role in mitigating its own harms—by plowing through mountains of data about weather changes, extreme weather events and human displacement. “The only way to make sense of this data is using machine learning methods,” Oliver says.
Milner believes that the best way to expose AI-caused systemic inequalities is through people's stories. “In these last five years, so much of our work [at D4BL] has been creating new datasets, new data tools, bringing the data to life. To show the harms but also to continue to reclaim it as a tool for social change and for political change.” This change, she adds, will depend on whose hands it is in.