In the early hours of Saturday 11 January, Prof Teresa Lambe was woken up by the ping of her email. The information she had been waiting for had just arrived in her inbox: the genetic code for a new coronavirus, shared worldwide by scientists in China.
She got to work straight away, still in her pyjamas, and was glued to her laptop for the next 48 hours. "My family didn't see me very much that weekend, but I think that set the tone for the rest of the year," she says.
By Monday morning, she had it: the template for a new experimental coronavirus vaccine. The first death from the new virus was reported around the same time, but it was still a month before the disease it causes was named Covid-19.
Lambe's team at Oxford University's Jenner Institute, led by Prof Sarah Gilbert, was always on the lookout for Disease X - the name given to the unknown infectious agent that could trigger the next pandemic. They had already used their experimental vaccine system against malaria and flu and, crucially, against another type of coronavirus, Middle Eastern Respiratory Syndrome (Mers). So they were confident it could work again.
The checks did not fail, and after seven weeks they had enough doses of the vaccine to start the first trial. By this stage 10,000 volunteers had been recruited - all of them signing up within a matter of hours. Elisa Granato, a microbiologist, and Edward O'Neill, a cancer researcher, were the first two to receive Oxford's trial jab in April.
For Lambe, the vaccination of the first individuals was a monumental step. "I remember walking home from the lab and almost breaking down because we had a vaccine and got it into somebody's arm," she says.
The team waited 48 hours to check Granato and O'Neill had had no serious adverse reactions. With both doing well, they immunised a further six volunteers, and then gradually ramped up the numbers.
But only a few days later there were shocking reports on social media that the first volunteer, Granato, had died.
It wasn't true. I spoke to her that day, and she told me she had never felt better. I
posted a bit of the interview online. It helped quash the story, but not completely. Bizarrely, there were suggestions the interview might be fake and she really needed to hold up that day's newspaper to the camera as "proof of life".
"It was very upsetting," says Gilbert. "A lot of work had to be done, for her sake as well as for the trial's sake, to make it clear that this was just fake news."
Disinformation spreads fast and has the potential to undo months of hard work. Green found it very stressful. "Oh God it's awful! To think how malicious that can be, how people are really spreading nasty things… that doesn't feel fair because my team were all doing their best," she says.
As the trial grew, it was clear that Oxford's small manufacturing facility would not be able to keep up with demand. The team decided to outsource some of the manufacturing to Italy. But when the first batch was ready, there was a snag - the Europe-wide lockdown meant there were no flights to airlift it from Rome.
"Eventually we chartered a plane to bring 500 doses of vaccine because it was the only way we could get it here in time," says Green.
This is a really important part of the story which ended up being highly significant months later.
The Italian manufacturers used a different technique to Oxford to check the concentration of the vaccine - effectively how many viral particles are floating in each dose. When the Oxford scientists used their method, it appeared that the Italian vaccine was double strength. What to do?
Calls were made to the medical regulators. It was agreed that volunteers should be given a half measure of the vaccine, on the basis that it was likely to equate to something more like a regular dose. This was partly a safety issue - they preferred to give them too little rather than too much.
But after a week, the scientists became aware that something unusual was going on. The volunteers were getting none of the usual side-effects - such as sore arms or fever. About 1,300 volunteers had only received a half-dose of the vaccine, rather than a full one. The independent regulators said the trial should continue and that the half-dose group could remain in the study.
The Oxford team bristle at any suggestion that there was a mistake, error, call it what you will. Perhaps the most accurate characterisation is that the volunteers were inadvertently given a lower dose. In months to come, they would be the stellar group in terms of vaccine efficacy.
From the start, the team at Oxford had had the goal of creating a vaccine that could help the world. To do that they would need billions of doses - something only industry could provide.
In November, two other vaccine trials published their results. They were astonishing. First Pfizer-BioNTech, then a week later Moderna, announced its vaccines were about 95% effective. The team in Oxford were encouraged.
Until then there was uncertainty that any vaccine could work on Covid-19, says Lambe. "There are so many infectious diseases out there that we can't impact."
Finally, on 21 November, the independent data safety committee was ready to reveal the Oxford-AstraZeneca findings. But the results were surprising - and more complex than expected.
Whereas Pfizer and Moderna had one efficacy figure from one big trial each, the Oxford jab ended up with three numbers: 70% overall, with the two full doses giving 62% protection while the smaller group, who were given that initial half dose had the highest protection, had 90%. It was a result no-one had expected.
Crucially, no-one who got the vaccine was hospitalised or got seriously ill due to Covid. Whereas in the control group there were 10 serious cases and one death.
The fact that those who had unwittingly been given the initial half dose from Italy showed stronger protection is intriguing. It may be caused by the immune system being primed more gradually, but the scientists can't yet explain it. Also, all the volunteers in this group were under 55.