The human stories of sequencing in lockdown

By December 2020, when Alpha was discovered, COG-UK centres had been working for nearly nine months. Together they had managed to sequence a total of 188,629 sequences from 2,406,290 infected people (7.8%) in the UK (COG-UK Dec 2021). By this point the consortium was sequencing around 8,000 to 10,000 SARS-CoV-2 genomes every week and looking to double this number in the coming months (Peacock Dec 2020). On 22 January 2021 Peacock reported that the Consortium had managed to sequence more than 200,000 SARS-CoV-2 genomes (Peacock Jan 2021). As of that month the UK had submitted 45% of all the SARS-CoV-2 sequences uploaded to the international GISAID database, far outstripping other countries (Cryanoski). The number of genomes sequenced by COG-UK continued to climb steadily upwards, reaching 250,000 by February 2021 (COG-UK Timeline). By October 2021, the UK had uploaded over one million genome sequences to GISAID, which accounted for a quarter of all sequences in the database. These had been sequenced by COG-UK in partnership with the UKHSA (UKHSA).

Figure 12.1: Tweet from Dr Sam Nicholls who helped update the CLIMB database for COG-UK work.

The high numbers of SARS-CoV-2 genome sequences could not have been achieved without the total dedication of many different people working very long hours in the laboratory and outside to contribute to the effort. This section delves into the reasons why people got involved in COG-UK and goes behind the scenes to see how it affected their daily lives and thinking for the future. This section tells the human stories underpinning the sequencing operation that is often just presented as dry, abstract, numbers.

Motivation for getting involved

What stands out when exploring the factors that drove people to participate in COG-UK's sequencing effort was scientific curiosity. Many jumped at the opportunity because they saw it as a way to contribute to their skills to help society deal with the pandemic. Dr Cristina Ariani, who was based at the Sanger Institute, for example, remembers 'I just felt I was lucky enough to have the job that I had, and to be able to contribute, and also keep my mind busy, not to be worrying about things. Although I was, of course, worrying, at least I was doing something, where I didn't just passively hear the news and see how things were going' (Ariani transcript).

For many members, COG-UK was a natural extension to their own research interests. This is succinctly put by Dr Sarah Buchan, who helped from Bournemouth University. One of the reasons she says she got involved was 'Probably just because, as a researcher, my gut instinct is always, if it's possible to generate data, you should. And it's clinical data, and therefore, it's really valuable. If I'd have thought about it for longer, which I suspect I hadn't at that point, then maybe Variants of Concern, immune escape and vaccines would have cropped up into my brain, but I think I'm crediting myself for too much foresight if I suggest that I thought about it at the time, because I probably hadn't (Buchan transcript).

The emergency situation created by the pandemic was also a very important driver. Some idea of how this shaped the response is articulated by Dr Sam Robson who led the COG-UK effort at Portsmouth University. He recalls, 'We were very much of the mindset that if we were asked to do something, we would try our hardest to do it. If we were asked, 'Could you do 200 extra samples this week', we would look at what we needed to do 200 more extra samples, and that's kind of how we ended up scaling everything up so significantly' (Robson and Beckett transcript). Robson's 'can do' attitude is also reflected in the words of Dr Andrew Jermy who was a senior advisor to COG-UK's communication team. From his perspective, 'At times it definitely felt like being thrown in the deep end. I think most people, we just jumped on board and said, we'll do what we can' (Harrison and Jermy transcript).

One of the strongest impressions that comes across from the interviews is the goodwill that drove COG-UK forward. Dr Michael Chapman, who helped set up COG-UK, argues that one of the reasons it was successful was because it managed to mobilise 'a coalition of the willing... Nobody was feeling coerced.' At the end of the day, he says, it was about finding 'enthusiasts' and harnessing their energy to move things forward (Chapman transcript). The same spirit is echoed by Ariani. She points out 'Everybody was under extreme pressure, but at the same time, everybody understood everybody was under extreme pressure. So people were trying just to be on it, and making sure that things were done in the most effective way, in the most collaborative way as well. It was quite amazing how much we've achieved with so many labs and hospitals as a group, as the COG-UK, in such a short period of time, logistically' (Ariani transcript).

What is also noticeable is how many people dived into helping COG-UK without really thinking about what they would get out of it. This is highlighted by the experiences of Dr Estée Török who was working alongside a number of people in Cambridge. According to her 'we were very lucky to have assembled a fantastic team locally that could just do this, and really wanted to work together, and got on really well. It was a really challenging time, but it was a really fulfilling time I think for everybody. I ended up working with people I had never worked with before, but it worked really well. There were no egos, we were pretty much self contained in a way. We didn't have to deal with the kind of politics and things that sometimes you have to do in research; we just rolled up our sleeves, got on with it. We weren't really thinking about academic careers or rewards; you were just thinking, this is what we can do, this is what we should do. If it's helpful, great. And I think that was the only thing that motivated us really' (Török transcript).

The same was true for Dr Leigh Jackson who came on board from the University of Exeter. He recalls, 'You could immediately feel that people were genuinely trying to help each other, to improve and get better, and do the best everyone could. Very rarely did people say "no" to things.' If asked to take an extra 500 samples this week, he says 'most people would say "yes" and find a way, rather than what I might be traditionally used to in certain academic circles, of 'no, unless there's money.' That was quite heartening to see and why I was interested in getting involved. I wanted to help as best I can, and it was good to see that there were so many like-minded people that were really in this for the right reasons' (Jackson and Thomas-McEwan transcript).

In many ways this 'can do attitude' reflected Professor Sharon Peacock's personality, which emanated through the consortium. For her, one of the ingredients that helped drive COG-UK's success was 'First of all, we were there to try and serve the country. And in times of crisis, you can't think about yourself, and I think that people bought into that vision and would have said that themselves. Working in times of crisis you must contribute what you can.' She also believes that what facilitated the enterprise was the people who got involved at the start. As she reflects, 'There was very little in the way of very big egos. I think that egos can really get in the way of effective consortium working. People left any sense of ego at the door.' She continues, 'The other thing that probably really helped is that most of us know each other. It's quite a small community and we know, like and respect each other' (Peacock transcript).

Just how much Peacock's leadership helped motivate people in the consortium was picked up by Dr Anna Kinsey, the Head of Epidemic Preparedness at the UK Medical Research Council, who sat on the COG-UK Steering Committee. What impressed her were the periodic thank you emails she received from Peacock, which went to all the consortium members, in which she acknowledged 'the difficulties and what hard work people have put in'. For Kinsey this showed Peacock's very positive leadership. She also remembered one of Peacock's emails at Christmas 2020, 'in which she asked people who'd receive it to also pass on thanks to their families who she knew had obviously had to make sacrifices so that they could contribute [to] the work that they had done. I think that's an important acknowledgement' (Kinsey transcript).

Juggling daily life with COG-UK work

Figure 12.2: Tweet put out by Sonia Crestpac showing toilet paper stockpiling in a supermarket, 3 March 2020.

What is easy to forget now that the emergency of the pandemic has begun to recede, is how much of a turbulent time everyone faced at the beginning. Dr Cordelia Langford who helped lead the COG-UK work at the Sanger Institute, for example, remembers some of the anxiety she felt 'hearing that people were panic buying while I was worried about millions of samples coming in and us being able to sequence them.' The pressure for her was taking decisions while learning that 'in supermarkets there's no toilet paper, there's no tins of tomatoes' (Langford transcript). Similarly, Rich Livett, also at the Sanger Institute, points out 'Right at the beginning, people's knowledge was very low about how infectious it was going to be. So people were frightened.' For him 'most of the challenge wasn't really around the work; it was the fact society changed enormously around us, and everyone was just expected to get on with it, which we all did. When I look back I think it is remarkable. You always reflect on people's experience of, maybe war, and think, how do people cope with that? But you don't have a choice, you just get up and get on with it' (Livett transcript).

One of the major challenges at the start was the uncertainty as to whether lockdown might be imposed in the UK. Some of the stress this caused is described by Dr Rachel Stanley and Lizzie Meadows. They were just about to launch a new laboratory information management system (LIMS) at Norfolk Research Biorepository, a site that subsequently handled all the metadata for COVID-19 samples sequenced by the Quadram Institute for COG-UK. The new LIMS had taken a year of planning and involved many different people, including all the biorepository team and third party systems suppliers. It was scheduled to go live at 9am on Monday 23 March. This turned out to be the day the UK Prime Minister ordered everyone to 'stay at home' and announced that lockdown measures would legally come into force three days later. Fortunately, Stanley and Meadows had planned for such a scenario, which included providing laptops and printing out paperwork so that people could work from home. But then everything was thrown into disarray when a number of their key team members came down with COVID-19 just days before the LIMS was due to go live. In order to rescue the situation, Meadows spent the weekend quickly getting on top of how to use Microsoft Teams so that she could coach everyone remotely to successfully deliver the project on time. Looking back she says, 'It was just incredible having this team that in the face of what was coming were all as determined as I was to get this over the finishing line' (Meadows transcript; Stanley transcript; Quadram Institute).

Figure 12.3: Newspapers announcing first UK lockdown. Announced on 23 March 2020, the first UK lockdown measures were in force from 26 March until June 2020 when the government began a phased reopening of schools in England and relaxed restrictions including the two metre social distancing rule. Thereafter a number of local lockdowns were imposed and a second national lockdown came into force in England on 5 November 2020. A third national lockdown was enforced from 6 January 2021 following the identification of Alpha (Institute Government).

Lockdown posed its own difficulties because many people had to shift quickly to working from home, which for many was a complete novelty. Some were better placed than others to cope with this. Chapman, for example, says his group sensed the 'direction of travel' before the announcement of lockdown so 'were mentally and emotionally prepared, as well as being physically prepared to work from home… People took their chairs and things like monitors home, to make sure they had a comfortable, safe set-up for themselves' (Chapman transcript). The communications team at Quadram Institute also acted early. Andrew Stronach, who headed it up, recalls, 'we didn't miss a beat. We all went to working from home quite quickly, before the government said we should. When you work in an institute with microbiologists and immunologists, we were probably a little bit ahead of that curve. And what we managed to do is move all of our communications online' (Stronach transcript).

But nonetheless, working from home took some adjustment, because, as Livett points out, 'people were trying to run their work lives from their personal space' (Livett transcript). Just how much juggling was involved is also depicted by Dr Julia Miskelly from Queen's University Belfast who recalls, 'We then all got sent to work from home and I remember getting my computer and everything set up because I had never worked from home before. I was in the bedroom, my husband downstairs, and then we had home schooling' (Miskelly transcript).

The closure of nurseries and schools over many months created its own strains. While some COG-UK participants were able to get places for their children in nurseries or schools because they were classified as key workers, others had to balance their work alongside performing childcare and home schooling (Abudahab transcript; Aigrain transcript; Nelson transcript). Dr Sonia Goncalves, from the Sanger Institute, remembers 'At one point we were all at home, we had our children at home as well, doing home schooling. All of that was adding to the complexity of the long hours that we had to work, because there was so much to do all the time' (Goncalves transcript).

Figure 12.4: Tweet from Professor Matthew Loose, who led the COG-UK effort at the University of Nottingham. His words capture some of the ways parents had to get their heads around how to home school their children.

Figure 12.5: Tweet from Loose providing a glimpse of the early days of parents trying to help their children at home. Blast stands for Basic Local Alignment Search Tool. Used by bioinformaticians, the tool proves a way to find regions of similarity between sequences.

Figure 12.6: Tweets from Loose reflect some of the challenges of working at home with children around.

Dr Ian Johnston points out that home schooling did not only affect people with children. Critically it had a knock-on effect on his management team at the Sanger Institute who said it required everyone needing to work flexibly seven days a week to 'fit around people's childcare arrangements at home'. He recalls that some of his team would be 'home schooling, and then getting kids to bed, and then we'd have catch-ups and calls and meetings at nine, ten o'clock at night when they were able to. Or very early in the morning. And then during the day we were doing the standard work as well' (Johnston transcript).

Figure 12.7: Tweet from Dr Andrew Jermy who stepped into the role of leading COG-UK's communications. His tweet captures some of the pressures people experienced on the home front as a result of lockdown.

Remote working could also be tough psychologically. Dr Alessandro Carabelli, who ran COG-UK's Mutation Analysis working group, says he sometimes 'felt like I was living like a monk. I tried to give myself a strong, strict routine. So, waking up at 6.30, and doing this, doing that, and, right, specific time. I was giving myself three hours to read the literature, and then, work on the coding for another amount of time. I think that probably helped keep me mentally sane' (Carabelli transcript). Similarly, Dr David Studholme, a bioinformatician at the University of Exeter, comments 'I was rarely out of my house'. He was just communicating through Zoom or Teams which he felt was not quite the same as having face-to-face meetings because 'you can't have quite the same bouncing of ideas and so on. We got better at it, but at first it was a barrier' (Studholme transcript).

Figure 12.8: Tweet from Loose illustrating how working from home became normalised over time.

Georgie McManus, who was remotely dealing with COG-UK's communications, says she missed 'a bit of the office chit chat and things that dispel the busyness a little bit. Obviously we do have lots of informal Zoom calls as well but I think it's a bit different being in the office with people every day' (McManus transcript). Conscious of the need to maintain a sense of community, the Quadram Institute ran online sessions called 'Random Coffee' which paired up people randomly to meet every week for a virtual chat. Stronach, who took part in them, remembers 'it was a really nice way to meet people that you wouldn't normally meet. It was purely a social thing. It was 'How are you doing? How's it going? What are you up to?' That worked really well' (Stronach transcript).

The transition to working from home was less of a shock for people like bioinformaticians. Dr Matthew Bashton, for example, from the University of Northumbria, points out 'I was already working from home, as a bioinformatician over the last 20 years, I've been involved in academic research in bioinformatics. Even when I was doing my PhD, I would log into computers from home when I was doing my PhD in Cambridge, so remote working wasn't anything new, I took to it very easily (Bashton transcript). Home working was more of an upheaval for laboratory-based people. This was the case with Beth Blane, who says 'Once I was fully involved with COG-UK and being completely computer-based, my days were spent on Zoom, on email or on phone calls, which was a big change to what I'm used to because I am laboratory-based' (Ludden and Blane transcript).

Figure 12.9: Tweet from Loose illustrating how much some people missed having the social interaction of meetings in person.

Some of the isolation imposed by lockdown was felt less by those able to continue working either in a hospital or onsite through lockdown. One of those able to do so included Professor Emma Thomson who was involved in the COG-UK effort in Glasgow. As she says, 'I felt quite lucky actually, almost reprieved, by being able to work in the hospital all through the outbreak because there's a good camaraderie' (Thomson transcript). The same feelings are expressed by Johnston who counted himself as 'fortunate' to be able to continue working physically at the Sanger Institute (Johnston transcript).

Working on-site, however, was not straightforward because of the rules around social distancing to protect everyone which limited how many people could come in. Then there were guidelines to consider around the wearing of masks and protective clothing. At first these were unclear and it took time to settle down, which added an extra layer of complexity (Nelson transcript). Eventually, Johnston states, 'You just get used to what two metres looks like, and stepping back from people, and having conversations, and as people pass, you know, moving around, it becomes second nature (Johnston transcript).

Workload

One of the key issues many COG-UK participants faced was how to manage their workloads. This was particularly heavy in the first few months. Professor Ian Goodfellow who was in Cambridge, recollects 'our workload went through the roof. We were in work seven days a week more or less, just working all the time' (Goodfellow transcript). Langford has similar recollections: 'I pretty much threw everything into this work. But that wasn't unusual. A lot of people did that.' She continues, 'It felt like regular working patterns just went out of the window. It was, all day, every day, working over weekends, but it never felt like a chore' (Langford transcript). The same sentiment is expressed by Tanya Brooklyn from the Sanger Institute who says, 'I was working horrendous hours, but that's what people did, because it was so important… I think this is probably true of most people who have worked in whatever area of COVID, whether it's somebody that's had to work in a factory, because they need the production lines to keep going to provide plastic ware, or food, or petrol pumps. People that I've come across certainly haven't grumbled about it. You are just united in working on this together, to get everybody through it. And although it's nothing like a war breaking out, there is that unity in times of trouble' (Brooklyn transcript). In a lot of cases, COG-UK participants indicate that in the first few months they were driven by adrenaline and the reward of contributing to the national effort (Thomson transcript; Watkins transcript; Aggarwal transcript).

Working at such a fast and intense pace, however, was not sustainable in the long run. John Sillitoe, from the Sanger Institute, explains, 'It's like a lot of those things isn't it, that you're so in it at the moment, and of course the nation is just focused on this one thing, you just get on with it. I think where it became more challenging was by the first summer, where you think you are running a sprint, but actually you realise you're running a marathon, except you're running a marathon at the pace of a sprint'. According to him, 'One of our biggest challenges actually was to get people to slow down, get people to take time off, to take holiday. Because everybody, in that summer, were still saying what we're doing is critical, it's important, we've still got lots to do, lots to get better. And, actually forcing people to take time off…. To slow back down again was actually really a hard thing… That was across the whole of COG, people were just so committed to what was being done. But we needed, for their own wellbeing, for their own health, to slow down, you know, you can only go at a certain pace for so long' (Sillitoe transcript).

Another factor that added to the strain was the nature of the work. This is captured by Amy Gaskin who was sequencing samples at Public Health Wales. She remembers 'Something I personally found quite challenging was trying to emotionally distance myself from the fact that each one of these samples comes from a virus that has ultimately infected a person, and trying to cope with the idea that if I'm uploading a thousand of these samples, that's a thousand people infected, and before vaccinations, before treatments or therapeutics, these people could be severely ill, and I have no idea if they weren't. So that was quite challenging for me to juggle that sense of responsibility. If I get this code wrong, or if I do something wrong, then that could impact the clinical output' (Gaskin transcript).

Staff capacity

What also affected the workload was the number and types of people available to meet the increasing demands of sequencing and analytics for COG-UK. At Portsmouth University, for example, Robson points out that they found it relatively easy at the start of lockdown to find wet lab staff as many who had finished their Master's degree were unable to find jobs. As he says, 'Having that ability to do something to tide you over through the pandemic, using your skills in a useful way to have an impact on the pandemic was quite appealing to a lot of people'. By contrast, they really struggled to find people with bioinformatics skills which left him as the only one who could do it. As he says, 'I was acutely conscious that if I got hit by a bus or something, or even if I got ill, that would have a massive impact on things because there was nobody else that could really run the samples through' (Robson and Beckett transcript).

The diagnostic laboratory at Barts Health NHS Trust also struggled to get sufficient staff. Dr Teresa Cutino-Moguel highlights 'we lost a lot of people in the lab, not only the people that were ill, but people that just left. We had a lot of vacancies that we couldn't fill so the lab was already understaffed. Then on top of that, during that wave, a lot of people were ill, or their children were ill, or their partners were ill. The point is that people were ill or isolating all over the place and we had loads of vacancies, so we were with skeletal staffing'. This shortage meant she and her colleagues had a 'really tough year' (Kele and Cutino transcript).

Figure 12.10: Photograph of Dr Teresa Cutino-Moguel. Credit: Bethany Lavin Photography. Born in Mexico, Cutino-Moguel decided to study medicine after learning about neuronal synapses at school. She says 'The small mechanisms that happen within our body and allow it to function or cause disease have always fascinated me.' When she studied medicine at the National University in Medicine in Mexico she had a chance to work in a laboratory of an immunologist whose enthusiasm and passion for research and to help others had an enormous influence over the direction of her studies thereafter. Following her medical degree, Dr Cutino-Moguel worked in a very under-resourced hospital in Mexico City during which time she witnessed the suffering of HIV patients. What was particularly heart-breaking was there were no drugs available in Mexico at that time. Cutino-Moguel subsequently moved to London to do a Master's in Molecular Biology Applied to Infectious Disease at the London School of Hygiene and Tropical Medicine and then completed a PhD at University College London in Immunology and Molecular Pathology focusing on intercellular HIV trafficking. Following a stint as a postdoctoral researcher at St George's Hospital, Cutino-Moguel was appointed Consultant Virologist at Barts Health Trust. Taking this position she says was 'an incredible opportunity, albeit very daunting as I was the only virology consultant at the time.' Describing herself as 'never one to shy away from a challenge', Cutiono-Moguel embraced the work of COG-UK as an opportunity to upskill the Barts team and scale up its sequencing capabilities (COG-UK Jan 2023).

Figure 12.11: Photograph of Shahiba Sultana. Credit Sultana. In 2016 Sultana completed a degree in Biomedical Sciences at the University of East London and then qualified to work as a biomedical scientist at Barts Health NHS Trust. Just before the pandemic Sultana helped streamline the procedures for testing of viruses at St Thomas' Hospital which proved invaluable for processing COVID-19 cases. Soon after this, in April 2020, she was appointed to become the senior biomedical scientist in the diagnostic laboratory at St Barts Health Trust. As soon as she moved over to Barts, she did a two week crash course in how to do next generation sequencing because her previous experience was with the older method of Sanger sequencing. This provided good preparation for when Barts joined the COG-UK effort. Sultana sequenced COVID-19 samples alongside routine samples for blood-borne viruses including HIV. Following her work in the laboratory during the day, Sultana would often go home to log on to her computer at night and weekends to study for the master's degree she was doing in Global Health at the University of Manchester. One of the modules she took for the course on pandemics brought home to her the value of what she was doing and her desire to pursue a career in global health. As she puts it, 'I just remember being like, 'oh my god, this is the dream job that I want, but I'm currently a student and I'm living the life working in a pandemic. And I'm studying at home about how to deal with pandemics, and I can't yet be part of the job I want to be part of (Sultana transcript).

At one point the Barts team had only three people to sequence the COVID-19 samples, which they needed to do alongside routine sequencing in-house for blood borne viruses, such as HIV, Hep B and Hep C. The team included Shahiba Sultana, who at the start of the pandemic was the only biomedical scientist on the team with any sequencing experience. A particularly intense time for her happened in April 2021 when she said she felt like she had 'samples in my face and I was sequencing, sequencing, sequencing so I never had time to stop and think'. Her attitude was very much about putting her head down and getting on with it. Describing the work as very manual, Sultana remembers 'Tuesday and Thursday would be the same steps, you're doing extraction, you're setting up the PCR. Monday, Wednesday, and Friday you might be doing another round of PCR and then you're going to sequence. And then Tuesday, Thursday, Monday, we're going to analyse it, running it through the pipeline, analysing your data and running back to the lab to carry on with the next step.' According to her, most of the days tended to blur into each other. On one day she calculated that she had pipetted over 1000 times before breaking for lunch. Not wanting to waste a single minute of her time, Sultana says that before she started work she already had a plan in her head of each step she needed to follow each day. What she says got her through the hard work, was her love of sequencing and the strong team support which was built up by the experience they went through together. Whenever one of them was down, she recalls 'we would just cheer each other up… We learned to just let everything roll off our shoulders and keep going because if we didn't laugh, we were just gonna cry' (Sultana transcript).

Laboratory work

The time-consuming repetitive manual nature of laboratory work outlined by Sultana was also familiar to other COG-UK participants. Angela Beckett and Robson, based at Portsmouth University, provides a detailed account in their joint interview of some the issues that cropped up early on, which got ironed out by putting in place checks, both at the sequence library preparation and downstream in the bioinformatics analysis, to prevent errors. Initially, the sequencing capacity in Portsmouth was restricted by the fact that they just had access to a MinION sequencer from Oxford Nanopore Technology, which meant they could only process 24 samples in one go, two of which were controls. One of the first steps involved hand picking out the positive samples provided in a 96 multiwell plate by the hospital to build up enough samples they needed to start sequencing. Beckett indicates that this could 'take the whole morning, hours sometimes' which she did together with Sharon Glaysher.

Before the hand picking could begin, Beckett and Glaysher prepared an experiment plan to determine where they were going to put samples received from different places on the plate. Beckett explains, 'Say Brighton sent us ten samples and Bournemouth sent us thirty, and somebody else sent us ten, we would have to take their sample list, arrange them into an experiment plan, and then we would tell ourselves where on our plate those samples would be.' The experiment plan meant that by the end of the process they would know the location of each sample on the plate. Once the laboratory work was completed the experiment plan was passed over to Robson so he could work out which barcode belonged to which sample (Robson and Beckett transcript).

The experiment plan was vital to making sure everything stayed in the same order during the pipetting process. This took a lot of concentration so that they did not accidentally put a sample in the wrong place. Beckett stresses, 'It's a lot of complexity and a lot of risk because when you're moving 96 things over they all have to be in exactly the same location each time.' One of the ways Beckett made sure she did not make a mistake was she would talk out loud as she was working, saying '1A to 1A, 2A to 2A, 3A to 3A, through the whole plate until I've done 96 to make sure. Because saying it out loud, especially in the beginning, it kind of connects with a different part of your brain, because if you start daydreaming, you'll be like, 'Oh, which one was I on again?' (Robson and Beckett transcript).

Having a 'freezer full of extraction plates' they first had to identify the right plate which could only be done by reading 'super long barcodes' that were 20 digits long. Once this was done, they then had to pick out the positive samples. For this, Beckett says Glaysher 'would read out the directions, like A1, and then I would repeat back to her just to confirm, “Okay, A1, I've heard you correctly”. Then she would say “Yes”, and then I'd have to cut the rubber bungs on the extraction plates because they come in strips. I would take tweezers, remove the bung, take the sample, replace the bung, and then I would sterilise the tweezers in between the samples, and we'd have to repeat that process' (Robson and Beckett transcript).

Once they had picked out the samples they then needed to carry out the steps of reverse transcription and amplification within the polymerase chain reaction (PCR) which also involved a lot of concentration. Using a single channel pipette, Beckett would take the sample and put it into a plate, add reagents to it and then once that was done she would have to 'move each sample one at a time over to the next plate that had the PCR reagents in it'. Because the 'PCR was split into two pools, so half the PCR was on one plate and half the PCR was on another' Beckett had to repeat that process twice. Summing it up, Beckett recalls ''I'd actually be moving anything from 24 to 96 samples by hand probably across four plates without making a mistake' (Robson and Beckett transcript).

Figure 12.12: Photograph showing microwell plates and tips box in Portsmouth. Credit Angela Beckett. Whenever Beckett started a new plate she would take a new box of plastic tips for the pipette, that come in sets of 96, and use them in the order she was pipetting. That meant, she says, ''If I got lost I could look at the tip box and I would know how many samples I'd done by how many tips were missing from the box that way I would know exactly where I was up to.' Another trick she remembers was 'if I was pipetting, transferring samples across into this plate, I would put tape over every column that I wasn't using, so I couldn't accidentally daydream and put it in the wrong one. It would take more time because I'd have to cover and uncover every single column strip on this plate, but it reduced the opportunity for mistakes to occur' (Robson and Beckett transcript).

Figure 12.13: Photograph of single channel (right) and eight channel (right) pipettes. Credit: Angela Beckett. Each pipette is operated with a button at the top which a person operates with their thumb to pick up and dispense liquid. In the beginning Beckett and Glaysher were picking out and transferring the samples using single channel pipettes because it took months for the eight channel pipettes to arrive, so for a long time Beckett says that they were 'stuck just doing everything individually'. The eight channel pipette made it possible to do eight samples all at once (Robson and Beckett transcript). Glaysher also points out that having access to the multi-channel pipettes meant they 'were less likely to make mistakes' and also 'definitely helped with the repetitive strain in the wrist and hands' (Glaysher transcript).

Figure 12.14: Photograph of eight samples being transferred between plates using an eight channel pipette. Credit: Angela Beckett. Describing the process, Beckett says 'If you imagine normally you'd have a plate with 96 wells and you would work top down, so you'd go 1, 2, 3, 4, 5, 6, 7, 8, and then up to 9, over and over and over again. But with multichannel pipette you can do all eight in one go so we're not having to handpick at this point.' Beckett points out that by the time the eight channel pipettes arrived, 'people were sending us plates with all positive samples so we weren't having to handpick positive samples from plates with a mix of positives and negatives anymore. We would have a full plate of all their samples, so it was literally a case of transferring from the plate they sent us to the next plate in the process' (Robson and Beckett transcript).

Figure 12.15: Photograph of plastic filter tip. Credit: Sam Robson. In order to protect samples from cross-contamination, each pipette needs to be fitted with a plastic filter tip which gets changed after each step in the process. Beckett explains, 'The tips are disposable, and they clip onto your pipette. You turn a dial on the pipette to set your desired volume. When you push down the plunger button, it displaces the air in the pipette so then when you release the plunger with your tip in the liquid-reagent it draws up the exact amount you need. The tips are designed so that at full capacity there is still some space at the top, so that you are unlikely to draw liquid into the pipette itself. To protect against accidentally drawing liquid into the pipette itself, you can also buy tips with a filter in the top, which is like a stopper. If you release the plunger too drastically, and the liquid shoots up, it's not going to get into your pipette, so you're not then cross-contaminating your samples' (Robson and Beckett transcript).

Figure 12.16: Photograph of samples being prepared by Robson. Credit: Bethany Lavin Photography. Each sample contains a number of copies of the original virus genome sequence, bound to a special adapter which is recognised by the sequencer and instigates the reading process. In addition, genetic material from each sample is bound to a unique barcode sequence, which provides a way of identifying each sample once the sequences have been read and interpreted by the computer. One of the advantages of the barcode is that multiple samples can be sequenced together in parallel in one flow cell, rather than having to sequence each sample individually. Once the data comes out the other end, a computer is used to separate out the samples.

In the first couple of weeks, Beckett says that she and Glaysher were relatively slow and sometimes they would take 10 hours to go through all the steps because they were new to the process and were 'triple checking everything'. Also, she remembers that 'some of the original protocols we were using wanted us to dilute each sample individually to balance the samples so that they were all the same concentration. In later iterations all that got condensed and we were able to skip some of those steps.' What also helped was the development of 96 barcode kits. Originally only 24 barcodes were available. The 96 barcodes were supplied in plates rather than individual screw up tops which also made the process faster. All of this was particularly helpful for when the Alpha wave happened, because as Beckett points out by then 'we were up- and running and everything was going smoothly' (Robson and Beckett transcript).

Figure 12.17. Credit: Angela Beckett. This figure is a collection of scanned pages from a lab book entry detailing one of the first times the Portsmouth team used the ARTIC protocol. The document highlights the 21 steps involved in running the ARTIC protocol to sequence samples. When discussing following the protocol, Beckett says at the beginning it took time to fully process. She compares it a bit like 'the first time you do a recipe when you're baking at home, you read the instructions once and then you go to do it, and then you're like, 'Oh, what did it say again?' and then you read it again, and again, and you want to make sure it's correct. So in the beginning it took a lot of time because you're triple checking yourself along the way. But I guess over time it becomes more automatic and it's much easier' (Robson and Beckett transcript).

The same intensity of work was experienced by both Stephanie Hutchings and Hannah Pymont at the Public Health Laboratory in Bristol which at the start of the pandemic was tasked to test everything in the local area. Not yet having the local diagnostic laboratories up and running in NHS laboratories, Pymont remembers 'we were doing an awful lot. I remember our record was something like 18 runs of the PCR unit of 96, plus a couple of control 96 well plates, in 24 hours.' Because the laboratory was so busy with that, Hutchings and Pymont were asked to help with sorting out samples to send to the Sanger Institute. Hutchings remembers this was a challenge because it involved a 'phenomenal amount of different things' (Hutchings and Pymont transcript).

In the beginning, when they first started, Pymont remembers they were doing manual PCR which was relatively straightforward because they were handling extracts which did not contain any live virus. But the movement to handling primary samples and automation actually created additional time-consuming steps to locate particular samples. At certain points Pymont and Hutchings were able to call on others to give them a hand, but Pymont says 'the reality was there wasn't always someone available' so it fell to them to get it done. Another issue Hutchings flags up was the fact that there were 'a lot of different protocols which meant that it was also difficult to provide the training to the other laboratory members.' She points out 'We tried, as best as we could, to automate and make things a bit easier along the way, but it was unpredictable. And with the surges in testing it was quite challenging.' Some idea of the intensity of the work is captured by Pymont's memory: 'we had our disco night in the CL3 [containment level 3 laboratory], where we had the disco music on. We got in at 8am and stayed until about 10.30pm just to clear some backlog, because it just wasn't going to happen otherwise. Like other people in the laboratory, you did what you had to do' (Hutchings and Pymont transcript).

Originally the Sanger Institute was performing the extractions on the samples sent by the laboratory in Bristol, but it subsequently did not have the capacity to continue doing this. So Hutchings and Pymont had to start picking out positive samples and put them on a multi-well plate to send to the Sanger Institute. Each of these samples then had to be mapped so the Sanger would know where they were on a plate. Pymont recalls, 'We had to assign them the anonymised COG-UK ID numbers. I spent an awful lot of time playing around with Excel and macros to try and make it a little bit more of a click-and-done kind of job, but to get to that process it took a long time. It was just quite fiddly making sure that these plate maps matched up to the coordinate, that A1 was where this sample was, and we were leaving spaces for things like controls as well, so it was making sure we left those in and then double checking as we went down. We had several situations where we were pipetting the things out following our plate map and realising we'd pipetted it into the wrong well, and then having to correct the plate map. It was something we all got very good at doing by the end of it, our pipetting is top notch these days after all that' (Hutchings and Pymont transcript).

Supply shortages

One of the running themes running through the interviews with COG-UK participants was the severe supply difficulties at the start of the pandemic. As can be seen above in the case of Portsmouth, this resulted in the delay of getting things like pipettes which had a knock-on effect on how many samples they could process. Getting hold of gloves and personal protective equipment was also an issue. Dr Rachel Nelson from the Sanger Institute points out that that was crucial because as she says, without that they would not have been able to operate (Nelson transcript).

Other items in short supply were also reagents such as ethanol, a basic sterilising agent, and lab grade water, both essential for laboratory work (Buchan transcript). To get round the reagent problem and shortages in PCR plates Cutino-Moguel remembers that at Barts 'We started acquiring different brands of platforms because the whole world wanted the same thing at the same time. There were reagent shortages all the time, or you would order something and it would be three months before you would get it' (Kele and Cutino transcript). Similar measures were taken by Dr Ana da Silva Felilpe and her group in Glasgow. As soon as they committed to the COG-UK effort and that they would see the number of COVID-19 cases increase they put in place 'parallel pipelines' which meant that 'if the supply of reagents was problematic for one pipeline' they could 'continue to sequence with the other pipeline' (da Silva Filipe transcript).

Figure 12.18: Tweet put out by Professor Ian Goodfellow, a year into the pandemic, illustrates just how much pressure there was getting hold of tips.

But what really sticks out for most people were the difficulties they had getting hold of plastic consumables, especially tips for putting on the end of pipettes. Just how many tips were required for COG-UK work is described by Beckett: 'Every time you're taking one sample from here to there, you would have to change tips. If you do 96 samples, you'd get through one box just on the first step, and then you go to the second step and you have to do everything again, so that's another box of 96 tips done, and then you transfer to the next bit, so that's another 96. I think we're getting through eight boxes of 96 tips per 96 sample libraries, a day basically (Robson and Beckett transcript).

Figure 12.19. Laboratories required a phenomenal amount of plasticware to do COG-UK work. These photographs (credit: Julia Miskelly) show how Queen's University Belfast managed to recycle some of the plastic boxes used to store tips into gate posts. The university has a research centre pioneering innovative manufacturing techniques to turn waste plastic into a wide variety of useful products (Martin).

The shortages meant that COG-UK participants were frequently having to spend what could have been valuable work done elsewhere troubleshooting to get over the problem. Aigrain explains how it affected things at the Sanger Institute: 'Normally we develop a pipeline using specific [plastic] consumables, we need them, we know we're going to use them and that we can't change them. But we couldn't find them anymore, so we had to constantly change consumables, revalidate the pipelines with those new consumables to make sure that it worked. It constantly added work that would have been unnecessary, but we had to be pragmatic' (Aigrain transcript).

Some of the pressure was eased by the fact that centres could call on COG-UK's network to help. At various points in the pandemic, Blane recalls Slack being used to ask around for items that might be in short supply. For example, she says, 'it would be such and such a site needs 10 micropipette tips, has anyone got any to spare? Or gloves? And another site would pipe up, 'yes, we've got extra', and they arranged to send them between them' (Ludden and Blane transcript).