daniel falush Profile picture
Statistical genetics of bacteria, humans etc. Hiring at all levels - do get in touch.
Oct 21, 2020 12 tweets 2 min read
I am a bit late to this party but this is what I have to say about "Great Barrington Declaration" about COVID strategy. gbdeclaration.org translated into 24 languages here. Two of its three authors, Sunetra Gupta and Jay Bhattacharya have separately published estimates of immunity through infection that were (a) scientifically flawed and (b) considerably higher than the truth but were most adept at getting public attention for these wrong estimates.
Jul 14, 2020 7 tweets 2 min read
Trump's ongoing battle of words with Fauci is a modern version of Canute's contest of wills with the sea. The worrying thing is if there is a sense in which Trump can actually win it. Such a victory would carve out a zone where facts have no import whatsoever. Fauci has as much scientific and for that matter moral authority behind him as is practically possible in today's world and is a savvy communicator. Furthermore, he almost literally stands up for facts.
Apr 9, 2020 14 tweets 3 min read
Thread. Will explain from first principles how China put itself in a good position to come out of lockdown, why it will be harder for the West and the broad principle it needs to adopt to make a smooth transition to normality feasible 1/15 The first principle is simple. You can come out of lockdown when (almost all of) the people who are infectious are separated from (almost all of) those who are susceptible, i.e. to those who have not been infected. 2/15
Apr 3, 2020 16 tweets 5 min read
Why the west needs to move towards centralised quarantining, as soon as is practical. Time-critical policy threat, please retweet. 1/14 @Magda_Skipper @genemodeller @mrjamesob @piersmorgan @guardian @Telegraph when, in early february, I heard about the plan to put coronavirus patients with mild symptoms into sportshalls, I felt uncomfortable for several reasons. I wondered about whether such facilities would provide a high standard of care. Wouldnt people do better at home? 2/14
Mar 29, 2020 16 tweets 3 min read
OK, bugger the weekend off, here is another politely angry thread on UK policy and how it *still* does not have the right urgency to it. 1/15 The very recent Imperial paper predicted as few as 5700 deaths, with a peak of 260 in one day. There were around 260 deaths *yesterday* and the numbers are still growing by around 30% a day. 2/15
Mar 22, 2020 12 tweets 3 min read
The "herd immunity" fiasco in UK policy making still matters in an immediate sense, because those involved it are still at the heart of the government response. So needs to be talked about now. Stress I have nothing personal against anyone and will not mention individuals. 1/12 The scientists involved just did not succeed in wrapping their heads around how serious a pathogen COVID-19 is and how it could not be business as usual any more. This is natural normal human psychology. Its really hard to break this news to a government or a country. 2/12
Mar 16, 2020 6 tweets 1 min read
Here is another way to understand how the scientists advising the UK government got their advice so wrong. All scientists are careful to say that COVID-19 is not like the flu, both because it is much deadlier and because noone has pre-existing immunity. 1/6 However the UK scientists basically tried to apply a flu-like model, emphasizing seasonality, immunity etc. Biologically, it is the most similar virus that kills many people. However, depending on circumstances, COVID-19 is approx 30 times more deadly than the flu, 2/6.
Mar 15, 2020 19 tweets 6 min read
OK, we need to be very very clear about what is going wrong with the policy in the UK. Boris is mortified not genocidal. #boristhebutcher is wide of the mark. 1/17 Secondly, the lack of transparency of the modellers who are advising the government is not deliberate. They are well-meaning individuals who are working very hard to predict the course of the disease and do the right thing. They have published plenty of good science. 2/17