COVID genomics on the Isle of Man: the history

Copyright Rachel Glover. If you want to use sections of this text ask first.

Get yourself a cup of tea, this is a long post and could take a few minutes to read. It’s been rather a nice exercise in pulling together lots of historical information, emails, papers and powerpoint presentations to aid my memory of what was a rather hectic time! It came about because earlier this evening I was asked on twitter about what I thought about IoMGov sending samples to Liverpool for genome sequencing. The reply was going to be far too long for a series of tweets, so here it is.

Needless to say I’m not going to sling any mud in this post, so if you’re thinking that you’ll get all the gory details of how badly I was treated by the DHSC for volunteering my expertise you won’t find them here. Saying that, when the Isle of Man Government inevitably pats themselves on the back for such an amazing genomics achievement at some point in the future, spare a fraction of a second for my blood, sweat and tears that brought them to the point where they even knew such techniques existed.

If you want a quick potted overview of what a genome is and how we can use them then download this presentation. I gave the keynote at this year’s UCM Festival of Research on 17th September and the download is a pdf of my slides. There might be a recording somewhere on the UCM website (I think).

Basically, genome sequencing is the only way to tell the difference between all the different lineages of COVID, including the “new” UK variant. The PCR test used to detect the virus in infected people can only say “detected/not detected”. Genome sequencing is super-high resolution testing that can, effectively, tell who gave it to who.

There’s a tiny bit of very basic science at the end of this post showing a small part of the IoM genomic analysis I did (and can get away with putting here) to make reading to the end worth your while 🙂

For many years I sequenced and analysed viral genomes for the UK Government (in addition to designing tests for them). I was at the cutting edge of viral genomics at the time (with 454 for those of you who know what that is) but if you want to take a look at small section of the work I did, take a look at my publication record. Our team was carrying out applied sequencing of ssRNA viruses routinely from 2009 before the clinical crowd looked at its real-world application in about 2013-14 during the Ebola outbreak in West Africa. So viral genome sequencing is my rather specialised bread and butter.

Back in February this year I did a Manx Radio interview on Women in Science day (I’m a LoveTech mentor) and at one point managed to get into conversation that COVID19 viral genomes were being sequenced worldwide to track the spread of the disease. Skip to 3 minutes on this link to hear me talk about how genomic epidemiology can be used for COVID, amongst other things.

A few weeks later I sent the now-infamous email that would change the course of my 2020 completely. It was an email to Steve Doyle (the head of the path lab at Nobles Hospital) offering to set up on-Island testing for them so that samples didn’t need to be sent to Public Health England in Manchester any more. I could write a book about how I went about doing that, but I digress. During that first week in the path lab, while Steve was still figuring out the best way to pay me for my time and expertise, I was making noises about sequencing the virus from whomever was unlucky enough to become the first positive patient identified on the Island.

On 22nd March I sent out the email below with this rather informative document (worth a download and was the first of many in those couple of weeks during setting up the lab) to Steve Doyle, Rizwan Khan and Rebbeca Shields at the DHSC. I was pretty sure we would get a positive patient in the next 3-5 days. The first positive case on the Island was identified two days later when the result came back from PHE from the swab taken from a patient on the 17th March.

As you can see, I was rather keen to get sequencing viral genomes as soon as I had on-Island testing up and running. Genome sequencing was always present on the agenda of those update documents.

At this point, not only was I carrying around two genome sequencers in my rucksack every day, awaiting the opportunity to use them, but I had also transferred a batch of viral genome sequencing reagents to the hospital lab “for emergencies”. I had been in touch with some old colleagues from the UK (notably Nick Loman (@pathogenomenick) and David Studholme (@davidjstudholme)) as I knew they were heavily involved in the newly formed COVID-19 Genomics UK Consortium (aka COG-UK). The Isle of Man doesn’t get included in these projects due to our non-UK, non-EU academic funding status but I wanted the IoM to be part of it and wanted to find out how we could get involved.

On 31st March 2020 the first patient samples were tested on-Island but the launch of total on-Island testing didn’t happen until three weeks later on the 20th April (lots of reasons, mostly government procurement related that I might talk about in future). During this time I had to assess the current skill sets of the biomedical scientists (which varied), identify the immediate training needs, and move and set-up robotic equipment belonging to Taxa Genomics in the hospital lab to automate the process as far as possible so that the biomedical scientists didn’t have to try to learn advanced molecular techniques within days that could go wrong without automation. I was also fielding hundreds of emails from various IoM Government, DHSC, Department of Enterprise and Nobles hospital staff asking for advice on various COVID19 tests, should they buy this-one-or-that-one, what is this new technology etc. Often these emails came to me second or third hand, with my answer being repackaged as an answer from whomever asked me for my input. It annoyed me that I wasn’t being asked directly but in hindsight, I know now that there are plenty of ego’s in the DHSC who wanted to pass my knowledge and expertise off as their own. It’s also worth remembering that during this whole period I was also running Taxa Genomics at a point where our DNA testing sales tripled during lockdown (!). Sleep was elusive, never mind viral genomics.

In the middle of May I saw that another scientific acquaintance had published a rather awesome paper using COVID19 viral genomic epidemiology to track transmissions in Australia. The next time I saw Steve I showed him the paper and told him we should be doing this on the Isle of Man. I assumed he would bring it up with Rizwan and Henrietta (Director of Public Health) at their (unofficial) testing strategy meeting on the 25th. When I heard nothing back I tried another tack and emailed Henrietta directly, copying in Rizwan and Steve.

This email was received positively and with some excitement by Henrietta and the team at IoM Public Health as the genomic analysis of the first outbreak could inform any changes in policy needed for any future outbreak. It could also be used in “real-time” to inform decisions on the ground during any future outbreak. However, it went down like a lead balloon within the path lab. I was given what could only be described as an email bollocking for ignoring a “chain of command” that I didn’t know existed. Henrietta invited me along to a section of their testing strategy group on the 16th June to discuss further. Rizwan was particularly annoyed about me going straight to Henrietta so I offered him final authorship on any papers I wrote about COVID19 genomics as a sweetener.

On 19th June myself and Rizwan had a meeting with Al Darby and Kathryn Jackson from the genomics facility at Liverpool uni, who were carrying out the COVID19 sequencing for the COG-UK project on samples from PHE Manchester. I was pretty sure a bunch of the Isle of Man samples tested at PHE Manchester would have already been sequenced, and it turned out I was right. But this opened an ENORMOUS can of worms.

You see, the COG-UK project only had ethical approval to test and analyse samples from patients in the UK. They hadn’t realised that there could be samples already sequenced by them and in the public domain that were from other jurisidictions not covered by their ethical approval. Which meant that before we could access the publicly available (anonymised) COVID19 genomes from the Isle of Man patients we would need our own ethical approval from the IoM Cabinet Office / Public Health. Only then would I be able to start matching the genomes up to patients, their contact tracing data and subsequent antibody status (my master plan for the first genomic epidemiology paper to ever be produced by the Isle of Man).

That in itself seemed like a simple, if drawn out, process. Write the ethical R&D application where I promise to make sure that patient data is fully anonymised, stored securely and that no-one would be able to be identified from any published analysis etc. Get approval, start doing some science. Simple, right?

Not simple. The can of worms was opened further. No-one had ever considered the legality of samples being sent to UK reference laboratories before, even though the hospital had been doing it for decades. Apparently all Isle of Man samples sent to reference labs in the UK had been sent illegally (in theory) because the legislation did not exist to cover these scenarios. Then there was the matter of consent. Had patients consented to the virus sample they gave being tested further? At first it looked as though we would have to retroactively consent all ~300 patients to ask if it was OK for us to test the samples further. That would have been complicated given a number had sadly passed away from their infection. Thankfully saner heads prevailed and the expected push back from the Attorney General’s office was countered with prior examples. Isle of Man samples containing Salmonella spp. and Mycobacterium spp. had been whole genome sequenced for typing at reference laboratories in the UK on many previous occassions and it was argued that COVID19 was no different. Dr. Rebecca Rowley in IoM Public Health was an absolute star throughout this whole period (it took weeks) and arranged for the Caldicott Guardian to sign off the legal side of things so that I could finally submit the ethical approval application for consideration. Phew!

Cue a long wait while the ethics application was written, submitted, assessed, amended and finally approved in September 2020. In the mean time the microbiology lab chucked out a bunch of those precious positive samples that needed to be sequenced on-Island. I still don’t the how’s or why’s considering they had been fastidiously saving positive samples at my request for genomics since March. I’ll probably never know.

Once the approval was through I was able to get my teeth into some of the data available on the public databases. I had spent time with Madeleine Sayle in IoM public health making sure that they retained the contact tracing data and working out the data management strategies for tying everything up computationally to aid my analysis. There were a few frantic emails in September making sure the contact tracing data from the original outbreak wasn’t deleted by the Information Commissioner! The small amount of genomics work that I did before I resigned was presented “hot off the press” to MHKs and MLCs at Tynwald on the 27th October, having literally analysed it the day before. I didn’t know when I presented it that I would resign in utter frustration later that evening, but I’m glad it was seen by a few people before I left! When I recorded the public version of that talk and uploaded it to YouTube I omitted the slide with the preliminary genomic results just in case there were any confidentiality issues I hadn’t considered. Ultimately, though, the data was open access and on a public database in the form I present it below.

There were 34 COVID19 genomes available from patients on the Isle of Man at first, representing all positive samples from the 17th March 2020 to 26th March 2020. If you remember, the border was closed on the 27th March so this was a great little starter dataset. I can’t present much of my analysis here as I can’t identify any patients, nor show any of the information I know about the clusters now that I’m no longer working for the DHSC. But the part that I can show gives a rather nice little diagram we call a phylogenetic tree.

Each genome is identified with it’s anonymous identifier, then the lineage assigned, followed by the date the sample was taken.

I’m assuming 99% of you reading this aren’t scientists and so I should explain the image above. It’s basically a clustering of the virus genomes based on how similar they are genetically. Samples that cluster together are highly likely to represent transmission chains between contacts, families, households or social gatherings. I know what a few of those are but can’t divulge. What it does show is that there appear to have been potentially 16 introductions of the virus to the Island in the 10 days before the border was shut. Many of these introductions will have gone on to transmit further in the community. There may also have been introductions of the virus from people who were never tested. There’s so much more to this and the initial analysis I did but I can’t put it online!

I would have loved to link to the open data with some amazing visualisations (like these here) but in recent weeks the Isle of Man data has mysteriously disappeared from the public databases. I’ll let you draw your own conclusion on that one – let the conspiracies start!

It’s a sad situation where I wasn’t able to achieve one of the main things that would have made the Isle of Man look scientifically advanced especially given that my company has all the equipment, reagents and expertise to carry out on-Island COVID19 genome sequencing in real-time. But it is what it is.

COVID testing explainer 2: What is a false negative?

virus

I’ve seen a lot of comments on various Isle of Man facebook groups over the last few weeks where people are claiming their negative SARS-CoV-2 test result must be a “false negative”. So let’s delve into the testing and what “false negative” really means.

The first thing to know is that a SARS-CoV-2 PCR test result is reported as either:

  1. SARS-CoV-2 DETECTED
  2. SARS-CoV-2 NOT DETECTED

As you can see the lab doesn’t use the words “positive” or “negative”. When a PCR test is carried out, it either detects the virus or it doesn’t and that’s how the results are reported to your doctor.

This paper shows that infectiousness starts around 2.5 days before COVID-19 symptoms start and the peak viral load (the amount of virus) occurs about half a day before symptoms start. By the time you start to show symptoms (and therefore become eligible for a test through the IoM 111 line) there should be an awful lot of virus present at the back of your nose and throat for detection. Detectable (but possibly non-infectious) viral RNA also hangs around in there for at least two weeks and even up to a month. PCR testing helps to differentiate people with COVID-19 from people with colds and flu so that they can be contact traced and the spread of SARS-CoV-2 contained.

All PCR tests have something called a limit of detection which is the lowest number of copies of the virus that the test can detect. The test we use on the Isle of Man has a limit of detection of 10 copies of the virus, which is very sensitive.

np_swab

A possible weak point in SARS-CoV-2 testing is the swabbing. The nasopharyngeal swab that needs to be taken to collect the virus has to go really far back into the nose. It’s not very pleasant (but thankfully only takes a few seconds) and there’s the potential for a false negative result if the swab doesn’t collect enough cells. However, the testing we do on the Isle of Man actually takes this into consideration and tests the amount of human RNA present on the swab as well as the SARS-CoV-2 RNA. If the level of human RNA is below the threshold we established during our validation testing the result is reported as INSUFFICIENT SAMPLE and the patient is asked to have a new swab taken.

If we didn’t do the human RNA test then a small proportion of the low quality samples could be from patients with COVID-19 but could be reported as negative because the lack of material in the swab reduced the amount of virus below the limit of detection of the test. These would be false negatives and given the potential for a false negative person to spread the disease thinking they’ve had a negative test, we introduced the human RNA check on the Isle of Man to minimise the risk of false negatives as much as we possibly can.

So what does “false negative” really mean?

Well, there’s bit more to it than just that term and what most people assume it is. There are really four different terms to understand: true positive, true negative, false positive and false negative. 

positives-negatives

Let’s look at hypothetical Bob who was thought up by me and doesn’t exist in real life.

Bob has had a cough for a few months, sees the symptoms of COVID-19 on the news and decides to ring 111 for a test. He says the cough is new. The test result is “SARS-CoV-2 NOT DETECTED”. A few days later he gets infected with SARS-CoV-2 after a sneaky visit to his friend’s house during lockdown (bad Bob). At day 5 he becomes infectious and at day 7 he starts to show symptoms of COVID-19. He calls 111 for a second test and his result this time is “SARS-CoV-2 DETECTED”.

Bob assumes that his first test is a false negative and tells his mates on facebook that the test is rubbish.

But bad-boy Bob is wrong. His first test was a true negative and his second test was a true positive.

So no false negative for Bob.

As you can see, a real false negative is a patient who has COVID-19 disease but where the PCR test does not detect any SARS-CoV-2 virus (usually due to insufficient swabbing) rather than a true negative followed by a true positive. Despite the facebook rumours, real false negatives are actually pretty rare because of the human RNA test we carry out, the limit of detection and the known sensitivity of the test in patients with COVID-19 disease (99.4%).

So in conclusion, the likelihood is that if your test came back negative you had something else like a cold or the flu, not COVID-19.

COVID19 testing explainer 1: What is a virus?

virus

This post is intended as an explainer for the general public

To understand how the testing works we first need to understand what a virus is and how each species of virus differs.

Viruses can infect all types of life, not just humans. There are viruses that infect animals, plants, and even ones that infect bacteria. Viruses infect life forms, but they’re not considered to be either alive or dead themselves. They’re really quite clever little hijackers. They gain entry to a host cell and then use the cell’s own molecular machinery to make more copies of the virus using a genetic blueprint of instructions contained inside the virus. The cell then becomes a microscopic virus factory which ultimately bursts and releases all the new viruses it just made. The new viruses then go on to infect other cells as the infection progresses. The ultimate aim of a virus is to make as many copies of themselves as possible, and often cause chaos along the way which we call disease.

We classify organisms by splitting them into groups – a taxonomy – based on their characteristics. The groups get split into smaller and more specific sub-groups until all the similar organisms are grouped together. Virus taxonomy splits viruses by how they look (morphology), how they copy themselves, what kind of genetic material they have, which organisms they infect and even the type of disease they cause.

The first big split in virus taxonomy is whether the virus genetic material is DNA or RNA. You’ve probably heard of DNA because it’s the same language our own genetics are written in. It’s made of four letters: A, C, T and G. RNA is a similar molecule but the T is replaced with a U. DNA and RNA can be both double stranded or single stranded.

dna_vs_rna

SARS-CoV-2 (the proper name for the virus causing COVID-19 disease) is a single stranded RNA (ssRNA) virus. This type of virus tends to mutate quite quickly and rather than “strains” they tend to form something we call a quasispecies. The mutations accumulate over time as the virus infects more people and this means that we can track the infections using genomic epidemiology techniques by comparing the full genome sequence of the virus (the A, C, U and G letters) of all the different isolates. If you’d like to see more info on this have a look at the Next Strain website, it’s where the scientists are putting the genome data during the outbreak. I’ll put a post about genomic epidemiology on my to-do list.

The letters in the SARS-CoV-2 genome code for the instructions to make the proteins of the virus. This is comparable to having the blueprints to build a house. The virus hijacks the host cell and uses it’s molecular machinery to make more copies of itself. As an aside, this is why computer viruses got their name as they’re malicious code which likes to replicate itself. The RNA inside SARS-CoV-2 codes for a number of proteins which sit on the membrane of the virus as well as for enzymes which do the copying of the instructions themselves (the RNA-dependent RNA polymerase, or RdRp).

SARS-CoV-2

Once we know the sequence of letters in the viral genome we can start to design tests to detect it using those letters to our advantage.

It’s alive!

The Rachomics Blog has started!

I’ve been wanting to set up a blog for about, oh, 12 years or so. I was previously a scientist within the UK scientific civil service so blogs weren’t exactly something that were welcomed.

These days, I run a molecular diagnostics company (Taxa Genomics) and my collaborative work with the Isle of Man Department for Health and Social Care during #COVID19 has prompted me to look again at finding the time to do some science outreach. Plus, 93% of you said you’d read it. I’m going to hold you to that.

https://platform.twitter.com/widgets.js

Let’s face it, I love science and think that everyone else should too, given how much it impacts our daily lives (even if we don’t realise it some times).

I promise I won’t swear (too much).

//Rachel