There’s no good place to stand inside the Pandemic Response Lab’s sequencing facility in Queens, N.Y. Take a wrong step, and you’re blocking a gowned-up lab tech as she ferries a plate with the RNA from 384 Covid-19 test samples to a machine that runs a polymerase chain reaction (PCR) test, identifies positives, and extracts them with a robotic arm. A data scientist hurries by, bringing more RNA samples to a device that will convert them into DNA. In the corner are genomic sequencing machines that piece together chopped-up bits of that DNA to identify what form of SARS-CoV-2 a patient has. Name a mutation you’ve heard of, and this lab has seen it, as it takes part in the national and global efforts to sequence virus genomes, spot new and dangerous variations, and stem their spread.

Listen to this story

Things are moving in this world with mind-bending speed. Earlier this year, the main concern was alpha, a variant originally identified in the U.K., which seemed to spread more easily than the original “wild type” coronavirus. Toward the end of April, the Pandemic Response Lab (PRL) spotted the delta mutation, first identified in India and believed to be even more contagious than alpha. A variant like alpha or delta is distinguished by changes to its genetic code that have effectively altered the virus’s behavior. Delta has more than 20 differences in its genetic makeup, some of which may make it easier for the spike protein to bind to human cells and help the virus evade antibodies. Sequencing has shown that delta cases make up almost 70% of Covid positives in New York City. The CDC estimates the variant is responsible for 83% of cases nationwide.

In late June, PRL (pronounced “pearl”) started gearing up to screen samples for “delta-plus,” a mixture of mutations first identified in India and South Africa. “On a weekly basis these variant definitions can change,” says Henry Lee, director of genomics at PRL. When they do, software must be updated, samples must be reanalyzed, and authorities must figure out how to respond. It’s like watching a Covid wildfire: Find a new mutation, watch it spread, calibrate a response, find a newer mutation, and watch the process repeat, hoping the response will be more effective this time out.

Lee’s lab sends its sequencing data to New York City’s health department, then epidemiologists use it to study questions such as whether a case is a reinfection or whether it happened in a vaccinated person. In this manner, they assess how dangerous a mutation is – how fast it might spread, how severe it might be, how effective vaccines are against it. Then public-health officials can decide whether to enact safety measures such as masking and quarantining. Sequencing data can also contribute to the long-term fight to end the pandemic by helping pharmaceutical companies plan booster shots.

And yet for all of the value genomic sequencing provides, no one paid the Pandemic Response Lab to start doing it. PRL was formed as a Covid testing site with city funding last fall, then months went by as Lee and his colleagues waited for guidance on how positive samples would be sequenced. Jon Brennan-Badal, the 35-year-old chief executive officer of the company that set up PRL, Opentrons Labworks Inc., remembers being on a call with city officials, trying to work out how to get samples to another lab for this purpose. “The conversation bottomed out, because we were like, ‘Oh, it’s going to take a lot of money to make this happen,'” he recalls. Insurers wouldn’t reimburse for it, because sequencing rarely, if ever, influences the treatment of a given case. PRL decided to begin doing the work for free. “It’s a nonrational business decision if you just care about shareholder value,” Brennan-Badal says.

Opentrons had the luxury of being privately held and of having available capital and equipment. Its primary business is supplying the kinds of robotics in use at PRL and other medical facilities, which have been busy throughout the pandemic. Soon the lab will expand its operation to other cities. In March, Opentrons announced the launch of four national Covid testing hubs – in Los Angeles, New York, Seattle, and Washington, D.C. – to help get children back into the classroom. The company will charge $25 for a pooled test; those samples will be shipped back to the Queens lab for sequencing, with that part, again, on PRL’s dime. “The approach that I took with this is no one is willing to pay for it, but everyone is happy to have it,” Brennan-Badal says. “If you asked them, ‘Hey, would you pay an extra dollar for it?’ they’d be like, ‘Whoa, whoa, whoa.'”

Scientists and public-health officials across the country would love to see the kind of information Opentrons is producing gathered en masse and fed into a rapid, comprehensive national surveillance system. For much of the pandemic, though, sequencing has been piecemeal, spread between state health departments, academic labs, hospitals, nonprofits, and commercial unicorns such as PRL. The system is so scattered, it’s generous to even call it a system. Different types of labs have different priorities for the data. Some work quickly, some take longer. Some have struggled to obtain samples. Fuzzy regulations have caused delays and headaches all over.

In perhaps the most hopeful sign of progress in the U.S., the Biden administration in April announced it would dedicate $1.7 billion to developing sequencing capacity. But so far that hasn’t amounted to a sea change, despite scientists’ pleas that the system needs to be fixed in a matter of weeks – not months, and certainly not years. “There’s a total lack of coordination and a systematic plan,” says Eric Topol, director and founder of the California-based Scripps Research Translational Institute, which is home to one of the country’s biggest Covid sequencing operations. “I just don’t know why it’s not moving like it should be. This should be going at a very high velocity, and that’s not happening.”

Kenny Beckman, director of the University of Minnesota Genomics Center (UMGC), has a vision for the future of pandemic fighting in America. The Centers for Disease Control and Prevention (CDC), he says, should create a robust, user-friendly tracker that surveils virus mutations in something close to real time, much as the National Hurricane Center tracks storms. Scientists in public and private labs would feed in sequences, which would be compiled to create straightforward maps of virus activity that everyone from health experts to the curious public could consult to see what mutations are in their vicinity.

Beckman says he “doomscrolls” hurricane maps himself, even though he lives in St. Paul. He thinks the CDC could create something just as useful and compelling for Covid. The agency does have a tracking webpage showing the proportion of positive tests and their given variants, but the 10 regional classifications it uses are almost comically broad, lumping together places as far away from one another as Arizona and Guam. And the data was, as of mid-July, a month old; while the CDC does offer a projection of current mutation activity, it’s based on estimates. It isn’t clear how the tracker might help anyone make an informed decision.

“The CDC ought to be the national Covid preparedness center,” Beckman says. And not just Covid. Done right, the platform could be applied to all kinds of threats, from other viruses to antibiotic-resistant bugs to foodborne pathogens. “All of the technology for carrying out this kind of sequencing, this kind of genomics, this is all quintessentially American,” Beckman says. “This is the country that manufactured and pioneered the methods to do this kind of work. It’s appalling to me that we weren’t also the leaders to use it.”

No one knows this better than Beckman’s colleague Daryl Gohl, who runs the UMGC Innovation Lab. In March 2020, as the novel coronavirus was spreading, Gohl and his teammates leaped into action, working with the university’s medical school to develop a PCR test for it. The lab had also been working on next-generation sequencing technology, so the following month Gohl submitted a request for funding through the university and the Mayo Clinic to start looking for SARS-CoV-2 mutations. He was denied. “Surveillance sequencing was not a priority,” he says.

He tried crowdfunding, entering a statewide life sciences competition called Walleye Tank, a habitat-appropriate riff on the venture-funding reality-TV show Shark Tank. He’d won the Minnesota contest before, but this time he mustered only $1,395, enough to sequence maybe 30 samples. At the time, the state had more than 600 new Covid cases each day. “We had this nice method, but we had no money to do any sequencing,” Gohl says. He applied for funding from the CDC over the summer, but his proposal wasn’t reviewed because of an administrative error. He tried the university and Mayo Clinic once more and was denied again. Some other labs started using the procedure, but Gohl’s team couldn’t sequence much themselves.

Even one of the country’s best sequencing operations, two states over in Michigan, has struggled to meet the challenges posed by Covid. Since 2014, the health department there had been building a system that can track flu strains, tuberculosis outbreaks, and foodborne pathogens. Within weeks of discovering its first Covid case last March, Michigan was sequencing positive test samples. “We’ve been basically teaching other states what we do,” says Marty Soehnlen, director of the department’s infectious-disease division.

But Michigan faced major limitations. Some testing sites would send samples daily for sequencing, others might send them once a week; logistical issues like snowstorms could add delays. All that, plus the minimum of five days needed to do the sequencing, meant it could take up to two weeks before officials had the information they needed.

This January the state public-health lab discovered that the alpha variant had arrived. By one point in the spring, when Michigan had become a global Covid hot spot, it had the second-highest number of alpha cases in the country, behind only Florida, according to CDC data. (The high number in Michigan could also have been because the state was actively looking for the mutation.) “At the beginning we were hoping that that variant information would help us do more aggressive contact tracing and circle those wagons to control that spread,” says Annette Mercatante, the medical health officer of St. Clair County, which was among the state’s hardest-hit areas. Sequencing reports would trickle in, and her staff would drop everything to contact-trace people with mutated versions of the virus. But the genomic data simply wasn’t coming back fast enough. Eventually there were too many cases for her staff to keep up.

The most practical effect the mutation surveillance had for Mercatante was to give her talking points to persuade the public to take extra precautions. She successfully fought attempts by school districts to shorten quarantine times from 14 to 7 days. “For us it was kind of like a no-brainer. We’re not going to do this now,” she says. Other advice, including on social distancing, was frequently ignored. It was left to increased vaccination rates and possibly the warmer weather to slow alpha’s spread.

Soehnlen’s group is about to expand the state lab, using federal dollars they got last year. They’re also hoping to work out how to visualize sequencing data so local public-health teams can use it more effectively – those hurricane-style maps, perhaps. But, she says, “that’s not something that’s super easy to do in the immediate.”

That’s where the Biden administration, in theory, comes in. The $1.7 billion funding announcement in April committed $240 million to states and other jurisdictions, with a total of $1 billion being invested over time. A further $400 million has been earmarked to create six Centers of Excellence in Genomic Epidemiology – partnerships between state health departments and universities. A final $300 million will be put toward building a National Bioinformatics Infrastructure, which would provide a uniform way to share and analyze sequencing data.

The CDC says it’s been increasing sequencing capacity, contracting companies such as Labcorp and Quest Diagnostics Inc. to do the work. That’s helped the U.S. rise from 45th place in January to 36th in terms of percentage of reported cases sequenced and submitted to GISAID, the global mutation database. The country’s standing has also been boosted by funding from nonprofits, including one project – by the Chan Zuckerberg Biohub and Chan Zuckerberg Initiative, financed by Facebook Inc. founder Mark Zuckerberg and his wife, Priscilla Chan – that involved actually doing the sequencing. The CDC estimates the U.S. is now sequencing 10% of all Covid cases.

The CDC has its own priorities for the resulting data. “It’s important to state upfront what the objectives of sequencing are,” says Greg Armstrong, director of the agency’s Advanced Molecular Detection Program. “They’re different at the national level than at the state and local level.” The national goals are to find variants that evade vaccines, don’t respond to existing treatments, or are highly transmissible. This is the sort of information that helps pharma companies update their vaccines. It may also help the U.S. government allocate resources – in July, it announced it would send surge response teams to areas where delta is spreading rapidly.

Given the patchy start to Covid surveillance last year, some experts have expressed concern over whether the CDC can create a system that will help bring the pandemic to a true end. Early on, according to Rick Bright, senior vice president of pandemic prevention and response at the Rockefeller Foundation, the agency failed to grasp the urgency of sequencing. “There didn’t seem to be a champion for it. It seemed to be a side project or a hobby from a well-meaning or well-intended group in the CDC,” says Bright, who famously clashed with the Trump administration over its handling of the pandemic while he was director of the Biomedical Advanced Research and Development Authority. “I would say even today there’s not a full recognition of the value of this genomic data even at the CDC.”

Many in the field are critical of how long it takes the agency to share Covid data publicly and the level of detail it provides. When a variant of concern is found, for instance, the CDC publicizes only the state where the test was done, which doesn’t help public-health researchers track its spread. Jeremy Kamil, a virologist at Louisiana State University Health in Shreveport, points out that the agency takes weeks to provide data for diseases such as influenza, too. “They like to centralize and have control,” he says. “We have the technology to change this, but there’s this rigid momentum.” The result, according to Stacia Wyman, a senior scientist at the Innovative Genomics Institute at the University of California at Berkeley, is that the CDC is “increasing the numbers broadly but not necessarily increasing the usability and usefulness for public health.”

Jade Fulce, a spokeswoman for the CDC, says the agency can’t provide more granular data about Covid mutations because sequencing of positive tests isn’t mandatory; it uses geographically broad regions on its tracker because there isn’t sufficient information coming in from certain states or jurisdictions. Fulce adds that the CDC has also formed a consortium of scientists who are doing sequencing. The more robust National Bioinformatics Infrastructure is a multiyear project and just getting started, she says.

For sequencing to achieve truly widespread adoption, health insurers may need to get involved. Right now, insurance companies don’t pay for it, because the discovery of a mutation doesn’t alter a patient’s treatment plan. “It’s not approved as a medical health activity,” says James Crawford, senior vice president of laboratory services at Northwell Health, the largest health-care provider in New York state. “We have not bridged that gap.”

Where Covid is concerned, there are generally few therapies to begin with. Finding ones tailored to variants will be that much harder in the short term because federal regulations prevent labs from telling a doctor when a patient has a mutation unless the patient is participating in a study. But if that problem can be cracked, and if sequencing increases broadly, it could help researchers develop targeted interventions, which would give insurance companies a reason to cover it. And that would in turn result in a new wealth of data for doctors, scientists, public-health experts, and the companies developing treatments and vaccines. “We have to do more than generate data,” Crawford says. “We have to generate knowledge.”

Gohl, the University of Minnesota scientist and Walleye Tank emeritus, found out in May that he did, in the end, get some money for sequencing from the CDC. It will be enough to sequence 6,000 Covid samples. Even as overall case numbers are on the decline and public-health measures are being relaxed, concerns about mutations, especially delta, are rising. There are still many people who remain vulnerable to Covid – children, the immunocompromised, the vaccine-hesitant. Gohl’s lab will be sequencing positive samples not only from Minnesota but also from nearby states such as South Dakota and Iowa. Missouri and Arkansas, where cases are on the rise again, have reached out for help. “This makes me very glad that we were persistent,” he says. It’s critical public-health work, getting under way long after it should have begun.

The very long game for the U.S. will be working internationally to guard against new threats. As of early June only 14 countries, all of which have developed economies, were sequencing 5% or more of cases and sharing that information through global databases, according to the Rockefeller Foundation. The foundation is trying to fill some small part of the gap, investing more than $20 million for sequencing in India, sub-Saharan Africa, and the U.S. Variants of concern continue to incubate in places where vaccination is limited. Delta-plus was initially spotted in travelers from Nepal, where just 6.3% of the population has been immunized, according to the Bloomberg Vaccine Tracker.

If another dangerous variant arrives in the U.S., “we’re at great risk of missing it until it’s too late again,” Bright says. “We have resources now, we have the science and technology. We have amazing willpower in the university system and the private sector. All we have to do is take the money and enable these groups and get the information as quickly as possible.”