Meet people who warn the world about new exclusive variants

In March 2020, when the WHO declared a pandemic, the GISAID public sequence database contained 524 covido sequences. The following month, scientists raised 6,000 more. At the end of May, the total was more than 35,000. (In contrast, scientists around the world added 40,000 flu sequences to GISAID throughout 2019.)

“Without a name, forget about it; we can’t understand what others are saying,” says Anderson Brito, a doctoral student in genomic epidemiology at Yale School of Public Health who helps make the Pango effort.

As the number of hidden sequences increased, the researchers who tried to study them were forced to create completely new infrastructure and standards. The universal naming system has been one of the most important elements of this effort: without it, scientists would make an effort to talk to each other about how the offspring of the virus are traveling and changing – to express a question or, even more critically, to sound the alarm.

Where the pango came from

In April 2020, some renowned virologists from the UK and Australia he proposed a system of letters and numbers to name new lineages or branches of the family of cobids. It had a logic and a hierarchy, although the names created — like B.1.1.7 — were rather mouth-watering.

One of the authors of the article was Áine O’Toole, a doctoral student at the University of Edinburgh. Soon, he will become the main person who makes that classification and classification, eventually combing hundreds of thousands of sequences by hand.

He says, “Very early on, it was just me who was there to direct the sequences. That ended up being my job. I think I never understood the scale we wanted.”

He quickly began building software to assign new genomes to appropriate lineages. Soon after, another researcher, Emily Scher, a postdoctoral fellow, built a machine learning algorithm to speed things up even further.

They named the software Pangolin, which refers to the debate over the origin of hidden animals. (The whole system is known as Pango.)

The naming system, along with the software to implement it, became a worldwide key. Although the WHO has recently begun to use Greek letters for variants that seem to be of particular concern, such as delta, these nicknames are for the public and the media. Delta refers to a family that is actually growing, which scientists call Pango with more specific names: B.1.617.2, AY.1, AY.2, and AY.3.

“When alpha was created in the UK, Pango made it very easy for us to look for these genome mutations to see if we had that lineage in our country as well,” says Jolly. “Since then, Pango has been used as a basis for reporting and caring for Indian varieties.”

Since Pango offers a rational and orderly view of what would otherwise be chaos, scientists can always change the way they call viral strains – so that experts around the world can work with a shared dictionary. Brit says, “It will probably be the format we use to track other new viruses.”

Many of the basic tools for tracking genome cobids have been developed and maintained by early career scientists like O’Toole and Scher over the past year and a half. As the need for hidden global collaboration exploded, scientists were encouraged to support ad hoc infrastructure like Pango. Much of this work was received by young researchers with technology expertise in their 20s and 30s. They used informal open source networks and tools, which were free to use, and anyone could volunteer to add changes and improvements.

“Students at the forefront of new technologies tend to be undergraduates and doctoral students,” says Angie Hinrichs, a bioinformatician at UC Santa Cruz who joined the Pangolin project this year. For example, O’Toole and Scher work in the laboratory of Andrew Rambaut, a genomic epidemiologist who published the first hidden public sequences he received from Chinese scientists online. “They were perfectly positioned to provide these tools that became absolutely critical,” says Hinrichs.

Builds fast

It has not been easy. Throughout 2020, O’Toole assumed responsibility for identifying and naming new lineages. The university was closed, but he and another PhD student at Rambaut, Verity Hill, were granted permission to enter the office. The commute, a 40-minute walk from the apartment where he lived alone, gave him a sense of normalcy.

Every week, O’Toole downloaded all the hidden deposits from the GISAID database, which grew exponentially each time. Then, he looked for mutations that looked similar or had a strange appearance and were looking for genome groups that could be mislabeled.

In particular when he got stuck, Hill, Rambaut, and other members of the lab would attend to discuss the appointments. But the hard work fell on him.

Deciding when the offspring of the virus will deserve a new family name can be as much as science. It was a tedious process, he went through a number of unheard genomes and repeatedly asked: Is this a new variant of the covid?

“It was pretty tiring,” he says. “But he was always really humble. Imagine crossing 20,000 sequences from 100 different places in the world. I saw sequences of places I had never heard of. ”

As time went on, O’Toole struggled to sort and name the volume of new genomes.

As of June 2020, more than 57,000 sequences were stored in the GISAID database, and O’Tool classified them into 39 variants. In November 2020, a month after he had to submit his thesis, O’Toole made his final solo career through the data. It took him 10 days to study all the sequences, which was 200,000 by then. (Although Covide has obscured his research into other viruses, he does include a chapter on Pango in his thesis.)

Fortunately, Pango software has been created to collaborate, and others have grown. An online community — one that Jolly noticed when the variant was expanding to India — was created and grew. This year, O’Toole’s work has been much more generous. New lineages are most often named when epidemiologists around the world contact O’Toole and the rest of the team via Twitter, email, or GitHub.

“He’s more reactionary now,” O’Tool says. “If a team of researchers anywhere in the world is working on some data and they think they have identified a new lineage, they can apply.”

The flood of data has continued. Last spring, the team organized a “pangothon,” a sort of hackathon, classifying 800,000 sequences into about 1,200 lineages.

“We gave ourselves three solid days,” O’Tool says. “It took two weeks.”

Since then, the Pango team has hired a few more volunteers, such as UCSC researcher Hindriks and Yale researcher Brito, both of whom initially participated by adding two cents on Twitter and the GitHub page. Cambridge University postdoctoral fellow Chris Ruis has turned his attention to O’Toole to help clarify the delay in GitHub’s requests.

O’Tool recently asked them to formally join the organization within the newly created Pango Network Committee for the Nomination of Lineages, which discusses and makes decisions about the names of variants. Another committee that includes the head of the Rambaut lab makes high-level decisions.

“We have a website, and email isn’t my email,” O’Tool says. “It’s been formalized a lot, and I think that will help scale it.”

The future

Some edge cracks have begun to appear as the data grows. As of today, there are nearly 2.5 million hidden sequences in GISAID, which the Pango group has divided into 1,300 branches. Each branch corresponds to a variant. Eight of them are to be seen, according to the WHO.

Because so much has to be processed, the software has started to curl. Things are being mislabeled. They resemble many strains because the virus repeatedly develops the most advantageous mutations.

As a stopgap measure, the team has built new software to catch things that use a different sorting method and can be lost by Pango.