A very interesting study on the genomic diversity of #SARSCoV2 based on the reconstruction of the 'mutational history' of genomes bypassing the phylogenetic tree reconstruction step, which makes the approach scalable and fast. 1/ biorxiv.org/content/10.110…
The results suggest that the progenitor of all #SARSCoV2 lineages in circulation was three mutations away from the earliest available sequenced genome (Wuhan-Hu-1), and was first in circulation around late October to mid-November 2019.
A start of the #COVID19 pandemic in late October to mid-November is consistent with the more plausible 'non-genomic' evidence available , including the detection of #SARSCoV2 in wastewater in December in Northern Italy. 3/ ncbi.nlm.nih.gov/pmc/articles/P…
This article is irresponsible nonsense. The number of inaccuracies and errors are staggering, and the language utterly inadequate. The entire narrative is in variance with all the evidence available in the public domain at this stage. 1/
I appreciate that everyone, everywhere wants their private #SARSCoV2 variant, scarier and nastier than all others, to blame for their woes. Though, I'm sorry California, despite your track record as trailblazers, your variant (B.1.427/B.1.429) doesn't seem to make the cut.
B.1.427/B.1.429 do not qualify as 'lineages of concern'. They has been around a while and while their highest frequency is in the US (California for B.1.429), They are not exceptional in their distribution, or recent increase in frequency. 3/ cov-lineages.org/lineages/linea…
I felt it may be helpful to provide an update on the new #COVID19 lineages in the UK and in SA that might both be more contagious than any #SARSCoV2 in circulation until recently. Those comments should be considered preliminary as the evidence available to me is still sketchy. 1/
Both lineages have accumulated an unusually large number of mutations / deletions. A plausible explanation is a persistent infection in an immunocompromised patient. There is earlier evidence that such persistent infection leads to fast viral evolution. 2/ nejm.org/doi/full/10.10…
All the mutations /deletions carried by the two strains have been observed before in other lineages. For example, both carry N501Y in the spike protein, a mutation that was first observed in Brazil in April and has remained at low frequency until recently.
"#SarsCov2 has never been isolated" strikes me as arguably the most absurd of all the '#COVID19 myths', and this in the face of stiff competition on all sides of the debate. I'll answer one last time and then I give up ... 1/
Antibody-Dependent Enhancement (ADE) is a complex and poorly understood phenomenon, likely underpinned by different mechanisms, where binding of a virus to suboptimal antibodies enhances severity of the disease, rather than providing protection to the host. 1/
The aspect of ADE of most relevance in the context of #SARSCoV2 vaccination is sometime termed 'original antigenic sin' (OAS), a phenomenon where the clinical course of a disease is more severe in people having been previously immunised to the pathogen.
The textbook example for 'original antigenic sin' (OAS) is dengue. The viral population is comprised of 4-5 genetically different lineages. Prior exposure to one lineage of dengue tends to make subsequent infections by another lineage more severe. 3/ elifesciences.org/articles/42496
There have been claims that #COVID19 has acquired mutations leading to more transmissible strains. We formally tested whether this was the case using 15,000 #SARSCoV2 genomes from all over the world:
... and the answer is no, not at all!
We analysed all mutations in the #SARSCoV2 genome that are plausible candidates for adaptation to its novel human host with a novel approach comparing the relative number of descendants of all sister lineages with and without a specific mutation.
No single mutation having emerged recurrently to this day is significantly associated with increased transmission of #SARSCoV2. Instead, all those mutations are neutral or deleterious to the virus' transmission, some significantly so.
#COVID19 is a pandemic. There will be many additional deaths worldwide, whatever we do. These will arise from #SARSCoV2 infections, and from indirect negative effects on healthcare (e.g. disrupted cancer treatments).
We also have to brace ourselves for the resurgence of other infectious diseases after years of progress (e.g. yearly deaths due to malaria and TB are expected to double). Discontinuation of vaccination programmes won't help (e.g. polio may likely return).
There will be the additional harm due to the economic downturn and disruption. Such knock-on effects range from increased domestic violence and suicide, to a stark rise in hunger globally, optimistically estimated to a doubling from the current ~9 million deaths/year.
Our paper on bacterial plasmids, the primary vehicles for antibiotic resistance and virulence genes, is out in Nature Comms. There is a 'natural classification' for plasmids and we reconstruct exchanges of plasmid-borne genes within/between species.
Hospitals worldwide are facing an increasing incidence of hard-to-treat infections, due to the burden of antibiotic-resistant bacteria. The situation is expected to worsen further due to secondary bacterial infections caused by the COVID-19 pandemic.
The majority of hospital-associated bacteria become resistant to antibiotics by acquiring resistance genes located on plasmids. We know very little about exchanges of bacterial plasmids, and plasmid-borne resistance genes, within and between bacterial species.
Our paper on host range of zoonotic pathogens is out. It reports by far the largest manually compiled host‐pathogen association database, (2,595 bacteria/viruses infecting 2,656 vertebrate species). It is also the biggest vertebrate phylogeny.
The majority of bacteria and viruses are specialists infecting only a single host species, with bacteria having a higher proportion of specialists compared to viruses. Conversely, multi‐host viruses have a more restricted host range than multi‐host bacteria.
Some key factors for zoonotic potential are the same in viruses and bacteria (i.e. evolutionary relatedness between hosts and being vector‐borne). Host phylogenetic similarity is the primary factor for host‐switching in pathogens (i.e. primates are worse than bats for us!)
Infectious disease epidemiology is largely about dealing responsibly with 'unknowns'. I've worked on tuberculosis (TB) for ten years, which killed ~1.4M people last year and likely many more over the next years. I suspect we may know possibly more about #COVID19 than TB.
TB control programs have been effective at reducing the burden of the disease over recent years (same for malaria and some other infections). Though, their implementation is brutally unemotional, and their objective is to save years of life with limited resources.
I cannot remember much emotional outburst about a grannies dying from TB, pregnant women from malaria or babies from bacterial diarrhoea. I'm not claiming this is the way it should be, and had we shown a bit more compassion, the burden of these diseases may be lower by now.
I've seen many alarmist and incorrect claims about mutations in #SARSCoV2 . In this thread I will address some of the major misconceptions. Parts of the material I cover can be found expressed more formally in an article we published recently. tinyurl.com/yb4mxhes.
Mutations are random errors arising during replication of the genetic material of an organism. A possible analogy would be mistakes made by Medieval copyist monks who were making copies of the bible over centuries.
Most 'copying mistakes' both in texts and genomes are unlikely to alter the information, and can be considered as 'neutral typos'. Of those that change the meaning of a text/genome, few are expected to improve it. Those that do are likely to be retained in the future.
I've been thinking about 'solutions' to #COVID19 for three months. I first expressed my doubts there was a 'solution' on March 14 (below). I've since then listened to many of the proposed 'solutions' and feel even more convinced now there isn't any.
The absence of a 'solution' for #COVID19 should not necessarily come as a surprise. Complex problems rarely have simple solutions, and pretending otherwise is populism. In the words of Thomas Sowell; "There are no solutions only trade-offs".
Suppression of #COVID19 can be achieved locally, but it remains somewhat unclear to me whether this entails any major long term benefit. A good drug (still to be discovered) and effective vaccines would be most helpful but neither are likely to spell an end to the pandemic
I've done many media interviews as of recently. Some felt rewarding and others less so. The most difficult ones where those where I had the impression the interviewer wanted to me to produce very specific statements that would fit into a pre-prepared piece.
I respect journalists, appreciate the constraints they are working under and believe their role is absolutely essential to democracy. As such, I'm trying to be generous with my time and do not expect or demand any prior knowledge of the field I happen to work in.
I don't wish to lecture anyone, but I may have some advice to journalists on how they may get the most out of me. Please feel free to ask me any question, but listen to my answers. Your piece will likely be stronger if you get answers from me you may have not anticipated.
It's been now slightly over two months that I decided to dedicate some of my time to inform the public on #COVID19 and try to fight disinformation. It feels like a Sisyphean task at times as the nonsense is coming so hard and fast from all sides.
The blatant untruths are easy to counter. What I hadn't anticipated and I struggle with is 'scientific populism'. This is largely fuelled by a few academics with no prior experience in infectious disease epidemiology, who have established themselves as COVID19 'experts'.
The typical manifestation of #COVID19 scientific populism is a mix of sloganeering, foolish optimism, a can-do attitude, complete disregard for facts, a dose of righteousness and a propensity to blame anyone who doesn't buy in the latest in vogue delusion.
I felt it may worthwhile writing a mini thread about #COVID19 immunity following infection to provide some context to the WHO 'immunity passport' statement and its rapid clarification/retraction.
There is no uncontroversial evidence for reinfection by COVID-19 at this stage. We do not know yet for how long someone infected will be immunised against subsequent infection, though on average it should likely be for a year or possibly more.
Importantly, even if residual immunity is insufficient to protect from #COVID19 re-infection, it is still expected to be protective insofar a secondary infection should lead to far milder symptoms.
I unashamedly take pride of my team's scientific papers. I also have no problem with 'metrics-based' evaluation of the value and impact of science, if applied thoughtfully. Science is highly competitive, and it does a disservice to society if scientists are not evaluated.
No 'Metrics-based' evaluation of scientists is perfect but it remains the basis of the fairest systems, as long as differences between fields and personal circumstances are factored in. Any alternative is highly subjective and leads to nepotism and cronyism.
'Metrics-based' evaluation of scientists is transparent and offers the best chance for exceptional junior scientists with 'unconventional' backgrounds to break through.
You're right. The initially available - poor and limited - data for the 2009 H1N1 influenza pandemic strain pointed to a fatality rate of ~0.4%. This indeed was an overestimate (the actual number is somewhere around ~0.05-0.1%).
The 2009 H1N1 turned out to be less severe than had been anticipated, and the scientific advisors who made some of the pessimistic predictions at the time got a lot of flak and have been accused of crying wolf.
I wouldn't fully rule out that the relatively complacent initial response by some governments to #COVID19 might marginally have had something to do with scientific advisors being worried of coming across as systematically 'overreacting' to outbreaks/epidemics.
I'm seeing a lot of claims, including from colleagues, that the UK Gov intentionally tried to kill a fraction of the population with a nefarious #COVID19 'herd immunity' experiment. This is simply not true and such conspiracy theories are damaging for the fabric of society.
Errors may have been made in the UK and other countries in the early stages of the #COVID19 pandemic but this was an exceptionally challenging situation to deal with from the start. I suspect that anyone who claims otherwise is not in command of the facts and evidence.
It is understandable that emotions run high during a pandemic and it is in the human psyche to try to find culprits for traumatising events. However, the UK (and other countries) need to get together as a society at this stage and accusations of wilful murder do not help.
Under the assumption that pregnant women are representative of the wider NYC population for #COVID19 infection, ≥ 15.4% in NYC have been infected for far (some of the women may have been infected before and resolved the infection which would not be detectable by RT-PCR).
87.9% #COVID19 asymptomatic is higher than anything reported for other demographic groups so far, except children. There is previous anecdotal evidence that pregnant women are not at high risk of developing severe symptoms, and even that pregnancy might be protective.
I’m afraid there is a misunderstanding. I do not object to 'digital track & tracing' and I even gave high level advice on how to implement it. I also cannot see why you frame it as 'my right to privacy vs. 'your' risk of dying. (1/4)
Digital track and tracing does not have to infringe on privacy rights, if thoughtfully implemented. See for instance white paper below for a proposal that minimally intrudes on our privacy rights. (2/4)
What I strongly object to is that #COVID19 should make us abandon privacy rights. To quote Edward Snowden: “There is, simply, no way, to ignore privacy. Because a citizenry’s freedoms are interdependent, to surrender your own privacy is really to surrender everyone’s." (3/4)
This preprint has been around a while tinyurl.com/vj9hp8h. Now that it's been picked by the media, it can't be ignored anymore. The modelling is quite sophisticated but some of the assumptions and interpretations are questionable, to say the least. (1/5)
The work builds on the assumption that any excess in reported Flu-Like Symptoms must be due to COVID. This is in itself problematic as unmonitored epidemics of all sorts of viruses, bacteria and mycoplasma causing mild flu-like symptoms happen all the time. (2/5)
The authors do not consider the #COVID19 epidemic in the US started before mid-January, fine. Though, they seem to observe an excess in Flu-Like Symptoms (eFLSs) often peaking in 2019 already. I found no attempt in the paper to explain the earlier non-COVID19 eFLSs. (3/5)