Cookies   I display ads to cover the expenses. See the privacy policy for more information. You can keep or reject the ads.

Video thumbnail
Thank you for joining us for our hot topics.
Today, we're very lucky to have Professor Michael
Lynn from Stanford.
Michael received his AC summa cum laude in biochemistry
from Harvard.
And he followed that with a PhD in biological and biomedical
sciences, also at Harvard at the medical school.
He also received an MD degree from UCLA in 2004.
Dr. Lin is currently a professor at Stanford,
where he leads a research group developing
protein-based tools for imaging and controlling
biomedical processes.
He has a long string of awards.
And to me, most importantly, he is the author
of what I consider as the most viral slide deck
on COVID-19 in the recent days.
Michael, welcome.
Welcome to CSAIL.
Welcome to a MIT.
Welcome to the College of Computing at MIT.
We're eager to learn from you.
Please go ahead.
Oh, great.
Hello.
Well, thank you very much, Daniela.
Thanks, everyone, for taking your time to join.
It's my pleasure to share with you what I've learned myself
about the virus and the disease in the last few weeks.
So you know, we're living through unprecedented times.
We've seen nothing like this since perhaps
the 1918 flu epidemic.
And I think it's useful for us to all learn the facts
and try to spread knowledge about the disease.
So I'm gonna get started and share my PowerPoint window.
So some of you have seen this slide deck.
Some of you have not.
And I understand this slide deck,
I had made for my own lab meeting
to basically ease anxiety among my lab members
and use the facts to try to put together a plan of action.
And so, now I'm happy to share these facts with you as well.
And I'll try to explain--
I'll try not to get too deep into the biochemical details,
but hit the most important points along the way.
So as some of you have seen, this
is not a pretty PowerPoint.
I haven't had the time to make it really glossy.
It's not like a TED talk.
I'm not pushing a simple idea.
There's just a lot of facts.
But I think it's only when you understand
all the facts together that things make sense.
So we'll just get started with some very general numbers
to put some context into the numbers we're
going to get into shortly.
The total population of USA is close to 330,000.
I've been doing a lot of my calculations
based on numbers in California because it's just easier for us
to think about it locally.
That's about 1/8 of the USA, 40 million people.
So, you know, the major causes of death
for young people in the US are traffic accidents.
There's 30,000 per year in the USA, about 3,000
for California.
For older people, actually flu is the leading cause of death.
There are 40,000 deaths a year.
It's never really known how much,
but the range this year is estimated
between 22,000 and 55,000.
So out of that 40,000, about 5,000 in California.
So for a population of 40 million in California,
you can use this sort of number to get a general idea of how
lethal flu is.
If 25% of people got the flu virus and only 5,000 died,
that's an infection fatality rate or IFR of 0.05%.
It's probably actually more like 0.1%, that is more
like 12% of the population might get the flu virus in a given
year because we have flu shots.
So is everything OK?
You guys can see the slides, Daniella?
OK.
Yes, everything is going great.
Thank you very much.
Yes.
OK, great.
So we'll get right into some quick definitions.
The disease name is COVID-19.
I think we have all seen this term much more than we want.
Technically, it refers to the disease, not the virus.
In practice, the term "cases of COVID-19"
are used to refer to having a positive lab test,
whether or not you have symptoms.
For reasons that are partly political, partly
medical, if you just have the symptoms, you know,
you're not necessarily yet a COVID-19
case because you could have, for example, a bad case of flu.
So you tend to wait for the lab test results.
The WHO, the World Health Organization,
introduced the name COVID-19 after holding the world
in suspense for a few weeks, saying how important
it was to get the name right.
And then when they announced the name,
there was the name for the disease,
but there was no name for the virus.
And so everyone started using COVID-19
as the name for the virus because there was no other name
given by WHO for it.
So this has led to some confusion.
Some allowed people use the term COVID-19 to refer to the virus.
I mean, I guess that's fine in a popular lay point of view.
But technically, it actually makes a difference because when
we--
I'd like to talk about cases of the disease and infection
numbers based on the estimate of the virus.
And so I try to use a different name
for the virus when possible.
I mean, I personally think COVID-19 is a terrible name.
Because it stands for "coronavirus disease 2019."
And who says, if you want to talk about a virus,
you could call it the COVID-19 virus.
That would mean coronavirus disease
2019 virus, which is a mouthful for saying nothing.
It basically reveals that the disease name
has no informational value.
So previously, we had named diseases
by some sort of description as a disease.
So SARS was named SARS because it
stood for severe acute respiratory syndrome.
And actually this disease is very
similar to SARS, both in symptomology and in the agent
that causes it.
So the World Health Organization could have just named
this SARS 2 and that would have been
both accurate and descriptive.
So what is the name of the virus?
It was initially called 2019-nCOV.
That stands for 2019 new coronavirus.
And it was given by some infectious disease
organization.
This hasn't been picked up because it's a really hard term
to remember.
You don't usually think of numbers
when you think of naming things.
So I think it's also misleading because it
gives the impression that this virus is actually novel.
It's not particularly novel.
And I would say it's the least novel of the new viruses we've
isolated and sequenced.
The most interesting thing about it is that it's not that novel.
And that's because it's 96% identical to the virus
that causes SARS in 2003.
So in fact, the biological name for it,
the name for its sequence is SARS coronavirus 2,
or SARS-CoV-2.
So we'll use this name throughout my presentation.
And I like this name because it reminds us that SARS-CoV-2
is very similar to SARS-CoV-1.
And so a lot of the things that we've
learned about the coronavirus one, which came out in 2003,
it's almost 20 years ago, is very
relevant to predicting what would happen
with SARS-CoV-2 and COVID-19.
So I'll just say, if you haven't realized it yet,
that I'm not a big fan of how WHO has managed this crisis.
But WHO decided not to name the virus SARS-CoV-2
for precisely this reason, because it made
people think of SARS.
They wanted to obscure the relationship between the two
viruses.
They thought it would be unfair to call the virus SARS-CoV-2
and to call the disease SARS-2 because it would remind people
about SARS.
Actually, I think that would have been a good thing to do.
And as scientists, it's better to be
clear about what you're working with, rather than trying
to create false distinctions.
So I call SARS-CoV-2.
So what our coronavirus is, what our SARS-CoV-2,
I don't have a picture.
I mean, I think you've seen many EM images, many illustrations
of the virus.
Like other viruses, it looks like a little round
ball with little spikes sticking out of it.
From the molecular biology point of view, what's interesting,
important to know is it's fully RNA.
It doesn't go through a DNA phase
and that we're actually very familiar
with this class of viruses because they account for 10%
to 30% of the common cold.
And they're unusually stable for RNA viruses.
So we're used to hearing that RNA viruses mutate
very quickly.
RNA viruses will always be less stable than DNA viruses.
But among RNA viruses, these seem to be pretty stable.
For example, the coronavirus OC43 causes the common cold.
It was sequenced once in the 1960s.
It was sequenced from an isolate from the 1960s.
And it was sequenced again from an isolate in 2001.
And those only had two amino acid differences,
six nucleotide differences.
So that's pretty amazing.
As you've probably heard, there are many coronaviruses in bats.
And they seem to hop easily between species.
So they're quite adaptable to a variety of mammalian species.
So we had the Middle East Respiratory Syndrome,
MERS virus, that hopped from camels to humans.
The original SARS virus hop from bats, probably
to civets and then humans.
And then this one most likely came from bats, as well,
through some intermediate animal.
And there's a bat virus that is very closely
related to the SARS virus.
You can see that in the bottom red arrow.
And I made this point that the human cold virus
OC43 is very closely related to this mouse hepatitis virus.
Actually, what's not showing here
is there's also a cow virus that's
very closely related to humans.
So either humans or cows seem to have given laboratory mice
hepatitis.
So how might you kill this virus?
Everyone wants to know how do you kill it.
So it's an envelope virus.
It has a thin plasma membrane that's
only two molecules of lipid wide.
So that's actually really easy to kill.
Anything that would dissolve a lipid will kill the virus.
That includes soaps or detergents, solvents,
organic solvents like ethanol, isopropanol.
I like to use Windex which is a detergent.
And bleach will kill it because it just kills off proteins
and biological molecules.
So it's also sensitive to temperature.
I mean, well, I should say that the original SARS was
sensitive to temperature.
So the current SARS-2 is likely as well.
And there's some numbers that you can look at.
For example, on the bottom left, this
is a graph of how much tighter reduction you get
with different temperatures.
Oh, I left my temperature numbers off.
That's actually, I think, 37, 33, and 28.
I'll have to put that back on.
And then the survival of the virus depends on the surface.
So this is a paper that was recently published in the New
England Journal of Medicine.
On smooth surfaces like steel and plastic,
it survives for longer than on porous materials
like cardboard.
So that means on paper towels or napkins,
the survival should be like cardboard or lower.
But also the viruses will get trapped in these paper fibers.
So it's not easy to get virus out of a napkin, for example.
And then it can be killed by UV.
So the amount of UV in sunlight probably
is enough to kill SARS-CoV-2 by 90% after about three hours.
At the current--
Michael, could I ask you a question?
Yes.
So I heard that the virus can survive up to nine days.
The data here does not take us to nine days.
Can you clarify?
Yeah, the data here takes us to three days.
So most headlines will say virus can survive up to three days.
Yeah, thank you for the question.
You can see here on the stainless steel and plastic
that there's--
the dashed line is the detectable amount of virus.
So you could see that at 72 you still
are a little bit above detection.
And then at 96, it's undetectable.
So a lot of places, a lot of news headlines
will say three days.
That's true, that's a three logged reduction,
so that's 1,000 fold.
I think it's fair to use--
It depends what you want to use, 10-fold or 100-fold.
I use 10-fold because I think your transmission
risks are significantly lower once you've had a 10-fold drop.
So I have a follow-up question, if you don't mind.
Yes, please ask it.
So we know that there is this shortage of masks in hospitals.
So if we look at this data, does that
mean that if we save the masks for four days, then
they can be reused?
Uh, yes, correct.
That's right.
You know, they'll accumulate dust and other things,
but as long as you can breathe through them easily, in theory,
you can just set aside the mask and wait four days
and pick it up again, yes.
OK, so should I move on?
Yes, please.
All right, yeah, so now that we know how to kill the virus,
we might ask, how does SARS kill us?
So as many of you know, you get initial flu-like symptoms,
sore throat, cough, fever, and aches, shortness of breath.
And for many, in fact, probably most people,
they only have these symptoms.
The symptom onset is 48 days.
So it starts about a week after infection.
In severe cases, it will spread to the lungs and cause
pneumonia.
And that's shown in the top image.
So this can be diagnosed by CT.
There's a wide variety in symptom severity.
So about 50% are estimated to be asymptomatic.
That's based on data from the Diamond
Princess, the ship in Japan.
5% will require hospitalization and supplemental oxygen
due to difficulty breathing.
And that 5% is taken from sort of estimates
based on diagnosed cases of 10% elsewhere adjusted
for the 50% rate of asymptomatic presentations.
Michael, we have some questions, if I may interject.
So one question is, is the disease severity related
to the dose of virus exposure?
Yeah, that's a really good question.
I suspect it is.
I haven't seen any data.
Because, obviously, you can't dose people
with different amounts of viruses.
But I suspect it is, perhaps.
What might happen is if you get a lot of virus at once,
you'll have more immediate infection in the lungs.
And you'll just start off with a higher viral titre.
And then your immune system is behind in the race.
And so an example this is the doctor
who raised the alarm in Wuhan.
His name is Li Wenliang, I think.
And, yeah, it's funny.
It's such a sad case.
Every time I say his name, I start crying.
Because he was only 39 and he died.
So that was really unusual.
He was treating multiple patients with the virus.
So it probably was because he was breathing a lot of virus.
OK, a few more questions, and Michael.
So does one maintain immunity to the virus
if they go through it once?
Probably.
I mean, there's no data that suggests
that you lose immunity.
So if it's like any other virus, then it's
likely that you are immune.
I mean, SARS patients have proven to be immune.
So I think we should work with that assumption here.
Also, how long after the infection
will you test positive?
In other words, you are infected today, you show symptoms,
let's say, eight days from today.
At what point will you test positive?
Yes, you will probably test positive
around four days, five days.
Yeah, so these are great questions.
So what happens in the course of the disease
is the virus will land in your throat
and it'll start replicating.
And it's now a race between your immune system and the virus.
So the virus might double or it might-- well,
it's not doubling.
It probably has a burst size of several thousand,
meaning one cell can put out several thousand viruses.
It will do that over a maybe 18 to 24 hours, so about one day.
So the virus is replicating really quickly.
For you to get a positive PCR test,
you need to have a swab that gets enough virus that
creates enough nucleic acid in the reaction for the test
to detect it.
So my guess is that you will need
to have the virus for three to four days
before you have a positive test.
I will have one more question, and then we'll pause.
There are so many questions on the chat.
I will ask you one more question.
We'll come back to other questions
later so you can go on with your presentation.
And so the question is what is known about the disease
severity--
Hey, Jill, I was just reading a question and I lost it.
One second, sorry.
It might have been the question, what
is known about its disease severity
with respect to MHC alleles?
What is known about why some people are
more or less susceptible to catching the disease?
Yeah, right, that's a good question.
So some of it might have to do with MHC alleles, as described.
Although-- Right, so yeah, it could be that.
It could be differences in innate immunity.
So your body's first reaction to the virus
is to produce cytokines and then have
cells that are infected with the virus actually shut
down protein synthesis so that viruses
can't reproduce as quickly.
So it's kind of like a fire alarm, the first step.
That might differ between people.
You know, you wonder why kids don't have
severe effects from the virus.
So perhaps kids have a stronger innate immunity response.
And it's only a few days later that your T and B
cells are activated in the antibodies
are produced by your B cells in sufficient number
to block the spread of the virus.
So the fact that some people come through it
without any symptoms at all, like kids
and like 50% of the population in general, my guess is it
has to do with differences in innate immunity.
And then how effective you are in having
your adaptive immunity, so that's your T cells and B
cells, that specifically recognize the proteins
on the virus, how successful you are
in getting those working in clearing the virus may
have to do with MHC alleles.
It may have to do with how many white blood cells you have.
It certainly seems to be more of a problem as people get older.
Thank you.
We'll pause.
We'll come back to questions after we hear more of the talk.
Go ahead.
Sure.
Yeah, so when it gets really bad is when you
need mechanical ventilation.
And that's about half the cases that require hospitalization
begin with.
And then, you know, when people die of the disease,
it's due either to direct breakdown
of the alveoli in the lungs--
and there's a photo of that in the bottom right--
or you have this massive release of cytokines
by your immune cells.
Because now you've recruited billions of immune cells
to your lungs.
And they're all trying to fight the virus.
And they release a lot of cytokines, which actually
can cause low blood pressure and multi organ failure, especially
kidney failure.
So, you know, this is a slide deck that I had put on the web.
And I have links for all the facts, just
to make sure the facts are double checked
and for more information.
So if you go to my Twitter feed, which is MichaelZLin, all one
word, Z as in zebra, MichaelZLin,
It's a pinned tweet.
You can download the slide deck from there in the future
if you want to browse through any of these things.
So yeah, it doesn't make for a pretty presentation,
but I think it makes for a more informative document.
All right, so now a lot of text.
So what everyone wants to know is, what is the probability I
will get the disease?
Like, how many people have the disease.
And that's impossible to know from looking at the case
numbers.
So this is something where the news
will report the case numbers.
It's misleading.
The case numbers are always less than the infection numbers.
So if you want to know infection numbers,
I think we have to go by--
We have to estimate the probability of death
and work back from there.
So how do we know the probability of death
from infections?
Well, you could start with some good data.
So you could start with a place that
has done the most testing per capita, and that's South Korea.
So right now in South Korea--
well, what you can see in this graph in the middle
is the number of cases diagnosed in orange.
And the number of deaths is in red.
So it looks like there's a small number of deaths.
But actually, we know that deaths, on average,
lack diagnosis by two weeks, and this means infections
by three weeks.
It's actually two to three weeks or infections
by three to four weeks, with diagnosis
tends to happen around one week after infection
when the symptoms first start.
So if you look at the death and case curves,
the death curves indeed lag the case curves by about two weeks.
So this case curve, as you can see, is sort of flattening out.
It's reaching sort of a constant rate of about 100 per day.
But the death curve looks like this curve from three weeks ago
or two weeks ago, which is right around here.
So it's still sort of in this inflection point.
It doesn't yet flattened out.
That's the death curve I just mentioned.
So the current total deaths occur
from cases diagnosed two weeks earlier ago.
So if we have 75 total deaths last week and out
of 4,000 cases two weeks earlier,
that's a case fatality rate of about 1.9%.
But we know that infections have to be at least
greater than the number of diagnosed cases.
So that means the true infection fatality
rate is less than 1.9%.
And it just depends on what fraction of the infections
were diagnosed at, you know, two weeks ago.
So it's somewhere below 1.9%.
Can I ask you something about that, Michael?
Mm-hm.
I've been watching the Johns Hopkins real-time cases
website.
And they report that for the resolved cases,
the rate is more like 10%.
How should we think about that?
The closed cases are the cases that
ended either in the patients being fully recovered
or in death.
Yeah, well, I think they're just slow to resolve the cases.
If you have a severe case, it takes about a month
to fully recover free of symptoms.
And, you know, these resolution numbers,
I'm not sure how they're done.
They might be waiting for people to be free of symptoms.
But maybe they just don't follow up
in the large proportion of cases.
My guess is they're not following up.
It's just people are too busy dealing with new cases
to worry about what happened to the old ones.
So yeah, the case closure numbers,
resolution numbers have been low for all the countries.
So how else can we get this IFR, this all important infection
fatality rate.
So an analysis of Chinese data gave an IFR of 0.5%.
Some people doubt the reliability of that data.
I haven't looked at it myself.
But the analysis of the Diamond Princess,
that's the ship in Japan again, gave 1.2%
as the infection fatality rate.
Now, on the Diamond Princess, some patients
may have been helped with this new medication remdesivir.
But also those passengers were on the old side.
And we know old people are more impacted.
So I think that 1.2 is sort of a good estimate.
Also, you could look at population distribution.
So if you look at the number of people over 60,
there's more in Korea than in the US.
And there's more in the US than in China.
So I would guess that the IFR for the American population
is probably around 1%.
So an estimate based on age adjusted data from China
also gave a very similar number, 0.9%
in the range of 0.4 to 1.4.
So the range shows you how difficult
it is to really pick a number.
But let's just go with 1%.
All right, so we'll move to the next slide.
Oh, I need to click.
OK.
So as everyone knows, death is much more likely
for the older population.
So that's unfortunately the case.
It's not that, you know, the same is true with flu.
So it's not a surprising fact.
But you can see that, and like the flu,
it seems to be much worse.
Now, this again, I think, is actually
death rates out of total diagnosed cases,
without a two or three week time lag.
So this might be actually a little bit
of an underestimation, meaning that the denominator
cases might in actuality be smaller than is shown here.
So this makes the disease much more severe than the flu.
Michael, we have an interesting question.
Right.
Is there any reason beyond age distribution
why IFRs would be different across countries?
So we've heard that in the US, mostly younger people are
experiencing severe disease.
Yeah, some young people experience severe disease.
But I think it's still a small proportion out of the total
infected in the young people.
I haven't heard of good data for differences in the age
distribution of severity.
There certainly will be some country-to-country data.
In countries where older people are more likely to smoke,
they ought to be more impacted.
So maybe that's a little bit of a difference, like maybe
in South Korea older people have a smoking history,
whereas that's less likely in the US.
But I think we lack the data on that.
Also, can the virus mutate into being more deadly?
We've heard that maybe there are two strains
and there are worrisome numbers in France.
It's not clear if there are more strains.
What can you share about that?
Yes, it can mutate to be more deadly, in theory.
There's no particular reason to think
that it will get selected for that, though,
meaning, well, viruses actually don't want to kill
their hosts too quickly.
They'd rather the host stay healthy and just
be coughing a lot.
So there's no reason, in particular,
to be very afraid of that yet, right?
So meaning, the virus is known to be stable.
It's known to adapt to a bat host in a stable manner.
I don't think there's any reason that it would be selected
for increased fatality.
So right, so I made an estimate of how many new cases.
We don't have to go through the numbers.
But basically you take the current--
I split it up into weeks.
I go week by week.
I take the number of deaths in the last week
and I multiply by 100 to get a number of infections
that happened in the week ending four weeks ago.
And I do this at the end of each week to update the slide deck.
So this smooths out day-to-day differences
in the reporting of deaths.
And it also doesn't require any calculus.
It's basically a numerical analysis.
It's like doing Riemann sums with a time interval of a week.
So what you need to know is sort of the rate of new infections
each week.
So if there were, for example, in California,
there were 10 this week, 10 deaths this week.
That means there were 1,000 new cases four weeks ago.
With a doubling time of the infection of seven days--
somebody people have estimated six days.
I just used seven so we could sort of make the analysis
simpler.
With a doubling time of seven days, then from four weeks ago,
we're now 16-fold higher in the rate of new infections.
There are probably 16,000 new infections
in the last week in California, which is one in 2,500 people.
And then in the Bay Area, we're more impacted in other regions
of California.
So I just assume its one in 1,000 in the Bay Area.
Now, what can you do when, you know,
you're in Boston, which is also heavily impacted?
So it might be also something like one in 1,000 people just
got the virus in the last week.
So what does that mean?
Well, 50% of people will be asymptomatic.
So they'll never know.
And this is based on, again, the Diamond Princess data.
It's also based on some mathematical analyses
of the China data and of data from passengers on evacuation
flights back to the US.
So 50% people will never know.
The average incubation is seven days,
but the transmission may begin two days before symptoms.
So that means a large portion of people
will start transmitting the virus without knowing,
even if they do eventually get symptoms.
So that means you have a risk of one in 1,000
from direct personal contact, perhaps, and higher
than that from touching objects that are commonly touched,
which are called fomites.
Those are things like door handles.
And the risk of contracting it from touching fomites
is in proportion to the number of people touching
that between cleanings.
So you know, if 100 people touch that door
handle before you touch it, now that risk
is increased to 1 in 10.
So that means we have to take action
to reduce the risk of acquiring the virus, either directly
or from touching.
So what would happen if we did nothing?
So obviously, exponential growth will continue with a doubling
time of one week.
And you can do a quick, simple model as shown here.
This is very amateur model.
Most of you can do better models.
Again, I didn't want to do calculus,
so there are no equations.
It's just doubling.
And then we reach a point where we run out of people to infect.
And according to this model, that's in late May.
And California officials estimate 56%, you know,
total infections.
Angela Merkel famously said 70%.
So it's something between 50% and 70%.
At that point, the virus runs out of targets
and it's harder for the virus to find new people to infect.
Everyone else has been immune.
So that makes the top infection rate
about 20% for the last two weeks.
And then four weeks after that, we'll
have deaths that result from that 20% infection rate.
And that's huge.
So if you accumulated all the deaths, those numbers shown
here in California, you have to multiply by 8 to get the US.
That would be 2 million cumulative deaths.
Again, this is really just an amateur, back
of the envelope calculation.
But I think it's better than other back of the envelope
calculations that you've seen because it uses reported death
numbers and this estimated IFR.
Professional models say 100 million cumulative infections.
That's 30% or you could say 56% in the case of the one that
was done for California.
An estimated IFR of 0.5 as a conservative estimate.
So I think my estimate is not too far off.
It might be the worst case scenario would be the 60%
and the 2 million.
But it could be better.
And there's some nice sites for following case numbers
and fatality numbers if you want.
So compared to the flu, which had a cumulative infection
rate of 27% and an IFR of 2%, this
is probably similar in total numbers.
But the flu was famously devastating.
The disease itself is probably worse now than the flu.
Because back in 1918, they didn't have ventilators
or drugs or antibiotics.
So this is why we need to famously flatten the curve.
There are an estimated 5% will require hospitalization.
And 2.5% of those will require ICU and ventilation.
And the average hospital stay is three weeks.
So that means that in the last three weeks of May,
we'll get 10%, 20%, and 20% of the population infected.
And then that four weeks later in June,
will results in 1.6 million, 3.2 million and 3.2 million people
needing hospitalization.
And due to the three weeks that they
need to stay in the hospital, that's
eight million overlapping at the same time,
compared to one million hospital beds.
So that means we should slow down the doubling time from one
week to at least eight weeks.
Of course, the slower the better.
So that makes the next month critical.
And that's why California is shutting down
and Boston, I think, has shut down.
Because we need to know how far down we can get the R0, which
is the reproduction number or how much
we can increase the doubling time.
And due to the delay between infection and diagnosis
and hospitalization, it will take us several weeks
to know if it's really working.
And the hard deadline for this is really early May.
Actually, I should update the slide.
I think it's no longer May 16th.
I think-- let me go back.
I guess you could say it's around May 15
that it would still be bad.
But, you know, it won't require martial law yet.
But if you got to May 15th, you would
have to basically confine people to their homes
by threat of arrest because otherwise any further infection
would overflow our medical system.
So what can flatten the curve?
So I'm optimistic that weather will help.
It's not gonna help a lot, but it's gonna help a little bit.
So based on that graph from earlier, a 10 degree
rise in Fahrenheit will increase the doubling time,
will lower the amount of virus survival by two-fold.
So I think it should--
I would assume it would increase the doubling time two-fold.
So actually, I gotta correct myself.
A 10-degree change, I think, increases the amount
of death tenfold of the virus.
So I think it should increase the doubling
time at least two-fold.
And in this, you'll also get people outdoors,
so keeping them apart.
So that's why the flu is seasonal, for example.
So the goal is to reduce the reproductive number R0, which
is the amount of people infected by each patient to 1.0
or lower.
The current rate is estimated to be two.
And the sequential infection interval is about seven days,
you know, basically, starting from around when shedding
occurs at five days to when symptoms start at seven days.
So if one present infects two others in seven days,
that would explain the doubling time of one week.
So dropping R0 to zero would make the doubling time
infinite.
And drops in between one and two will
increase the doubling time in proportion to the dropping R0.
So you know, we have been able in other countries
to drop the R0 to 1.0, to a constant very small rate
of new cases in South Korea, Taiwan, Singapore,
and Hong Kong.
And that's actually without shutting down the economy.
That's by doing social distancing, very, very robust
testing and tracing and very strict hygiene, especially
the use of face masks and being very careful what you touch.
Michael, I wanted to follow up with a couple of questions
from the chat.
People are wondering, Wuhan was very successful in flattening
their curve.
Why are the rest of us struggling?
What should we be doing more?
Right, I think it's too early to say.
You know, in California we just implemented
the lockdown on Tuesday.
It will take--
And the cases are ramping up.
So a lot of the increase you see in the case numbers
is just due to testing.
And that's probably true even in places
like Italy where they've had a lockdown for a little bit
longer.
So yeah, you know, as a reminder,
I was to get infected today, I wouldn't show symptoms--
If I was to be infected today I wouldn't--
And if I were to show symptoms, I
wouldn't show symptoms for another five to seven days.
And currently there's a delay in the US testing.
So I probably wouldn't be tested for another seven days
after that.
So now there's this two week delay
between when infections start and when testing is complete.
So we won't see any change to the R0 or the doubling time
at least for another two weeks.
And in fact, from news reports, it
seems like most people with symptoms
are being told to stay at home to see whether they
get better or worse.
They're basically not testing unless they get worse.
And so that means a large proportion of mild cases
are being missed right now.
As testing expands, it expands to those mild cases
or expands people who are just household contacts.
And then that's gonna explode the number of diagnosed cases.
So that means we're not gonna see a reduction by looking
at the case numbers.
The only way we could see a reduction, I think,
is looking at the death numbers.
Because those will follow the infections
by three to four weeks.
But that's still a faster time window
than the ramp-up of the testing appears will be.
Michael, lots of other questions and activity on the chat.
Let's see, can you make a comment about face masks?
Yes, we're getting to it.
Everyone in Asia says you must use face masks.
It's the only way.
In the US, we're told face masks don't help at all.
Yes, I have four slides on face masks.
OK, we'll wait on that, then.
OK, let's see.
OK, one more question.
What is the best case scenario?
You told us the worst case scenario.
What is the best case scenario for recovery?
Yes, in terms of numbers, I assume?
Right.
I mean, the best case scenario for a person--
[INAUDIBLE] in numbers, yeah.
Yeah.
Right, so the best-case scenario for a person
is asymptomatic disease.
The best scenario for recovery without vaccines
is probably something like the next line
in the slide I was going to get to,
which is complete household isolation.
Like, for the country.
Michael, for the country.
Oh, for the country.
What's the best-- let's say everyone stays at home.
What's the best-case scenario?
How many more weeks to recovery from where we are?
Right, so I think if everyone stays at home,
it's likely we can limit the R0 to something like 1.0
on average, maybe even lower.
So in Wuhan, they did complete household isolation,
which is similar to what we're doing now in California,
except they probably have fewer exceptions
for vital industries.
So here in California, we have exceptions
for a whole slew of essential industries, including,
marijuana dispensaries.
But in Wuhan, they had complete house isolation.
And, you know, they took infected household members
to central quarantining locations or hospitals.
And they practiced strict hygiene.
And this dropped the R0 from 3.9 to 0.32.
And now Wuhan or China is reporting no new cases.
Some people doubt that's actually true.
But regardless, they have a large drop
in the number of cases.
So if we practiced complete house isolation, then,
you know, the disease would go away.
But that would require complete stoppage
of even essential industries.
So I don't think that's likely in the US.
But it's possible we can drop the R0 to something like one.
And then once better drugs or a vaccine comes along,
then it would be safe.
The big fear is pre-symptomatic or asymptomatic transmission.
In Wuhan, when they took the symptomatic patients out
of the home, the R0 fell from 1.3 to 0.32.
So this suggests that only 24%, or that fraction, 0.32
over 1.3, of transmission events occurred before symptoms.
But another study estimates that 44% of transmission
is in the pre-symptomatic period.
That's done by modeling.
So, you know, I'm not able to evaluate
those models or the quality of the data behind those models
so I don't know how good those numbers are.
But, you know, we can assume something like 24% to 44%,
or maybe a third of transmission events
occur before getting symptoms, without symptoms.
So how about completely asymptomatic patients?
So the undiagnosed carriers are estimated to be about 50%
as infectious as the diagnose cases.
The undiagnosed carriers in this study,
which was just published in Science this week,
include people who will eventually get the disease.
So that's not completely asymptomatic.
But undiagnosed carriers are more
likely be asymptomatic than diagnosed ones, obviously.
And 50% of people who get the disease
seem to not have any symptoms at all.
So this implies that asymptomatics are not
as infectious as symptomatic people.
So go to the next slide.
Yeah, so how effective is this?
We talked a bit already about the China,
South Korea, Taiwan, Singapore experiences.
But you can see that here on this graph.
The y-axis is log scale.
So that means a line would be exponential growth.
And so you could see that some countries are still
on a line, like France, like Japan, actually, and the US.
But others have become stable.
Again, this is case numbers.
So some of the initial growth in cases, especially in the US,
is just from faster testing.
So everyone wants to know if there are drugs.
So this is the part where I probably will go quickly.
Because this is mostly done for a biochemical audience.
But I'll say that there are a few drugs that
are very likely to work.
So one is this drug that's already approved
in Japan for pancreatitis.
It's an emitter of a class of enzymes, proteases,
that are required for entry of the virus.
And what's interesting is it was tested in cell culture.
And it has wide ability to block the replication
of all sorts of coronaviruses, including
SARS-CoV-1 and SARS-CoV-2.
And that's shown here in the bottom left.
It's also been tested against the original SARS
virus in mice.
So it's always nice to see things tested in mice.
Because the main barrier to a drug being effective
is getting high enough concentrations in vivo--
I mean, not having any toxicity in vivo.
Things that are toxic to humans often
show up in toxicity in mice.
So it's a good proxy for humans.
So you can see here that this drug camostat reduced
the amount of death from SARS-CoV-1 infection in mice,
compared to nothing at all.
So another one is this anti flu drug
called Avigan or favipiravir.
And this is a drug that gets converted
into a ribonucleoside, which is the building block of RNA.
And so then it gets used by the virus RNA polymerase.
And the virus is now unable to replicate its genome,
or at least not able to replicate it correctly.
The idea is it creates so many mutations during replication
that the virus eventually fails to replicate.
So this, again, works against the current virus
in human cells.
And then it actually seems to have really poor potency.
We use a concentration required for a 50% inhibition
as a benchmark for the potency.
This number is actually really high.
You usually want it to be below one micromolar.
This is 62 micromolars.
But actually, you can load this drug
to a high dose in humans, 400 micromolar in your blood, which
is really high, with just a pill.
So this is really efficient.
And it was declared two days ago to be
effective against the virus in controlled trials
by Chinese authorities.
No data was provided, so we don't really
know how good the data was.
But it sounds promising.
This one Kaletra, down here, is an HIV protease inhibitor.
It was thought that maybe because the SARS virus also
uses a protease, a protease is an enzyme that
cleaves to other proteins.
And it's required to process one big protein in the virus
to many different parts, individual parts.
Sorry, excuse me.
It was thought that maybe this HIV protease inhibitor
would be useful.
And there was anecdotal experience that it was useful.
But it turned out, just announced two days ago--
and I haven't had time to update the slide,
and I will right after the presentation--
that the controlled trial failed.
And so that shows the importance of doing controlled trials
and not relying on anecdotal evidence.
And I always thought that this rationale was very weak.
So I'm not surprised that it failed.
Chloroquine is getting a lot of press today.
And I think it's long overdue press.
It again has shown to inhibit viral replication
of the current virus in cell culture.
And a whole month ago, the Chinese Health Ministry
said that it was superior to control treatment
in multiple trials in China.
But again, no data was shown.
And actually, I'm not sure why they don't show data when
they make these announcements.
It's a little strange to me.
I actually-- Yeah.
They're cynical reasons that they might not be doing it.
Or it could just be a cultural thing.
But in any case, they haven't shown the data,
but they made the announcement that it was useful.
And then just two days ago, basically, the same conclusion
was reached by a French group.
It showed that it was effective in reducing
viral titers in patients in an open-label trial.
But chloroquine and it's relative hydroxychloroquine,
they're both similar.
They're both anti malarial drugs.
They do have some side effects which can be serious.
But they have been safe to use for preventing malaria.
You know, I think, you know, it's funny when the press says,
oh, it's not proven.
Yeah, maybe we shouldn't use it yet.
I'm like, well, you don't have to wait for FDA approval.
It's already a drug that you give for malaria prevention.
The current disease is a lot worse
than malaria in some respects.
So I don't see why we wouldn't just go on and use it.
Another drug that gets a lot of press is remdesivir.
And that's because it's probably the most
specific of the current antivirals for SARS-CoV-2.
So remdesivir, It's another prodrug
that gets converted into a ribonucleoside
and is used by the RNA polymerase.
It was originally designed to inhibit the RNA polymerase
for another RNA virus, Ebola.
But it was shown in a published paper
to inhibit the SARS-CoV-1 RNA polymerase and the virus
replication in mice very effectively.
And it also works in human cells on the current virus.
And that's not surprising because the two viruses
are 96% identical at the amino acid level.
So for most purposes, you could think of the SARS-CoV 1 and 2
biochemically as essentially the same.
And so now it's being tested in humans in multiple trials
in the US and in China.
The China trial data are due in a few weeks.
And now it's gotten a lot of good anecdotal results.
But again we should not rely on that.
That's why we have the controlled trials.
So this is a lot of text.
But basically, I should just note,
that they are drugs just to treat the symptoms and not
the virus.
So it does this often from cytokine release syndrome,
as we mentioned.
And there are drugs for that.
And so China has approved one of them,
which is a Roche antibody against a cytokine to prevent
cytokine release syndrome.
There's poor evidence, but some evidence that COX-2
inhibitors--
that's just Vioxx and Celebrex which were previously
prescribed for arthritis--
can prevent cytokine release syndrome.
But no one's really brought that up recently.
Again, like, treating this is good,
but it's not as good as preventing
the virus in the first place.
There's this big raging debate going
on in the medical community about ACE inhibitors.
And I'm not going to cover it.
Because I think it's really besides the point.
You know, internal medicine doctors
love playing with hypertension drugs.
And ACE inhibitors are hypertension drugs.
You know, some of you are probably
on three hypertension drugs.
I mean, this is just--
internal medicine doctors live for doing this stuff.
And they're fighting with each other
over whether having more or less ACE2 activity is
good for the virus.
I think it's a side show.
They're talking about preventing lung injury
or decreasing number of receptors for the virus.
I think it doesn't matter.
I'm not even going to touch on it.
So I'm not even gonna cover it.
But anyway, there's a link here for more clinical trials,
if you're interested.
So there's a whole bunch of new things.
You can make antibodies from patients
or you could screen antibodies in libraries in vitro
and just produce antibodies that will bind to the virus.
So this has already been announced by two companies Vir
and Regeneron.
I'm sure there are many more companies doing this.
This is a tried and true method.
This is why sometimes you get a travel shot of immunoglobulin
before you go traveling for different viruses.
So it can be quite effective.
And so, you can also make more drugs
based on the ones that already know, like that flu drug,
like the remdesivir.
And you can screen for drugs, either in vitro
or virtually on a computer.
So all these things are being done.
And then, of course, what we really want
is a vaccine that we can give to the entire population.
And the fastest right now are these RNA-based vaccines
where they just inject RNA into your blood.
And some of it goes inside your cells
and your cells will produce viral proteins.
And you'll enlist an immune response
to those viral proteins.
So there's some more--
there's a nice-- for chemists, there's
this review of all these different approaches.
I wonder why we don't just fix the virus with formaldehyde
and inject it, which is how the first polio
vaccine was made by Jonas Salk.
And a fixed SARS one vaccine was found
to be protective in monkeys.
And this has been proposed for SARS-CoV-2.
It's not like nobody's thinking about this.
So I'm guessing it's been done in China.
It's been proposed in this article that was
written by a Chinese scientist.
So I'm sure it's being done in China.
But you don't hear anything about this in the US.
You know, the CDC has its own lab.
And they used to do their own research.
I don't know if that's been cut recently.
But if they're still doing that, this
is what they should be doing, just making
the fastest, simplest vaccine and telling us
that they're doing it.
So that gets me to the next point, which is--
actually, I should probably skip the many failures
of the CDC and the FDA.
It's interesting.
You can read about it.
I need to look at the time.
Oh, we're out of time.
So let's skip this.
If you guys are interested--
Michael, we are still interested and we're still here,
those of us who can stay online will.
And you still have a few minutes.
We'd love to hear.
Yeah, we're gonna get to face masks, which is what was asked.
So here's a little table of the symptoms of COVID-19
versus the other thing.
So it's good, it's useful for knowing
how it's different from the cold or allergies.
But basically, you don't tend to have
the runny nose or sneezing.
But you have all the flu like symptoms of a fever, cough,
shortness of breath, headaches, aches.
So recommendations for hygiene, there are many recommendations.
But essentially, there are two types.
So the first type is being careful to keep the things you
own and your own hands clean.
So I'm a liberal user of hand sanitizer.
So the alcohol in hand sanitizer will kill the bacteria.
So you hear these things like, wash your hands often.
Don't touch your face.
Like, what does it mean?
How often is often?
Why can't I touch my face?
I'm gonna touch my face right now.
It's totally fine.
Somebody freaked out the other day when I did that.
I'm like, I'm at home.
Everything is clean.
So basically you should establish clean zones.
This is similar to like, if you ever done
surgery on an animal or watched surgery,
obviously you sanitize before surgery.
And then you have a clean zone.
Once everything is clean, clean things
can touch other clean things.
So that means that you should sanitize your hands
right after touching things that are touched by others.
You should preferably only use clean hands
to touch other people's things or handing things to others.
But clean your hands again afterwards.
And that means also sanitizing objects you get, right?
So like, if someone drops off something for you,
I would actually sanitize the outside of the package.
If it's cardboard, I wouldn't worry about it too much.
I would just leave the cardboard there.
But if they're giving you a smooth object
like, you know, yeah, like a CD, you know,
something that can hold the virus, I would wash it.
And sanitation means washing with soap
or using hand sanitizer or using Windex or using bleach.
And so again, if you already have
something clean in your house, you don't have worry.
You don't have to keep on rewashing it.
You don't have to keep on rewashing your hands
when you're at home.
Right, so people ask about disinfectants,
you know, like Lysol or things that have bleach.
These are made for large areas that you can't really soap
and wash or put alcohol on.
If you can use soap or alcohol, you
don't need to also bleach something.
Like, no one tells you to put your hands in bleach after you
wash them with soap and water.
So if you can wash an object with soap and water,
you don't need to quote/unquote disinfect it
with a disinfectant.
So I just keep my hands to myself as much as possible.
Face masks-- short answer is yes.
So the US are fond of saying that they're ineffective
or that you shouldn't buy them because caregivers need them.
But both of those statements cannot be true simultaneously.
If they were ineffective, then the caregivers
would not need them.
So they are effective.
So they block 95% of viruses if worn properly.
And that's because viruses are carried in droplets.
And basically, the masks are filters.
So, you know, there's a study done between surgical masks,
which is the simple, three-ply masks that look like they're
made out of paper and cloth.
And then there's the cup like masks, the N95 respirators.
The study said they're actually similarly effective.
But the N95 respirators are thought to be more effective.
So my recommendations for mask use
are based on relative risks.
The reason we've been told not to use them
is to save them for the caregivers.
And because the risk that you'll run into somebody who's
emitting virus in public is very low, so basically one in 1,000
at the moment.
So if you're just walking by yourself in public,
you'd be wasting a mask.
But I think they would be useful when you're
in proximity with strangers, so when you're
riding in an airplane, a train, in an Uber or Lyft
because you don't know if the driver may be contagious,
and especially if you're going to go to a hospital or clinic.
So--
What about the store?
Yeah, so it depends on how crowded the store is.
I avoid stores, so I don't know how crowded they are.
But I think if you're going to go in,
I would have a mask ready.
If you're going to go to a supermarket
and you see that you have to be waiting in line,
I would put on a mask.
And then if you're immunocompromised
or you're elderly, it's probably a good idea to put on a mask
any time you're in public, just in case
the guy who's jogging by you breathes on you.
So once the infection rates climb, then it's
useful for everybody in public.
And then they're absolutely recommended
if you're sick because you block the acceleration of viruses
from a sick person.
So again, the only controversy of masks
is due to its limited supply.
But I think people are wising up to this.
The former FDA commissioner Scott Gottlieb
just tweeted his support for mask use.
So I think you can only deny the facts for so long.
So this is sort of an example of public health officials
giving people partial information
that's deliberately misleading.
I think they've been disingenuous.
And it's caused people to lose trust in them.
You know, they should just be honest.
You know, people are adults.
If you tell them masks are effective but the rate
of infection right now is so low and we don't have
enough masks to go around that we'd like you not to buy them,
that makes complete sense.
But telling people you shouldn't use them because they're not
effective for you, it's true.
But it's true only in a very misleading sense.
And it doesn't engender trust when
people find out the full truth.
So you just need some filtration.
So if you don't have a mask and the guy next to you
is coughing, then just put your nose into your sleeve
and breathe through that.
I mean, I do that when I go to the bathroom anyway.
It's a really smelly bathroom.
And you know from experience that that blocks odors so it
must be filtering something.
And in fact, a study shows that a t-shirt blocks
70% of virus-sized particles.
So that makes sense.
So I won't go into this, how to use masks
and how to conserve masks because I
think most people just want to know whether it's a good idea
or not.
But I would say that in Asia, like in Taiwan and South Korea,
China, Japan, it's not just OK, it's
expected that you wear masks.
And in mainland China, it's required
that you wear masks in public.
So even if they have a very low infection rate,
they have enough masks to go around.
Yes?
Michael, for what it's worth, we're
trying to mobilize our alumni network to help us with masks.
Because I think you're right.
It's important.
Yeah, I mean, the US, you know, I mean,
the US is increasing its manufacturing of masks.
But it's still slow at the moment relative to need.
So the washable masks are not as good as the N95,
but they're still good?
Well, the surgical masks and N95 masks are equally unwashable.
There's cloth masks that you can buy, like allergy masks.
And again, they're probably better than nothing.
But they're not as good as surgical masks, which
are the three ply, you know, flexible ones
or the cup-like N95 masks.
But on this slide, I suggest decontaminating it
by heating it up in the oven at 70 degrees
Celsius for 30 minutes or putting them
in a UV sterilizer.
But as we mentioned before, if you just
let them sit in a clean place for a week,
the virus should be dead.
So I have a list of things that I think are OK to do,
like visiting relatives if you're really
careful with the hygiene steps.
But, you know, I would definitely
recommend buying groceries online
instead of in a grocery store.
You know, you can say, well, the person who picks and packs
your grocery might be infected.
But the same is true in a grocery store, the person who
stocks the items on the shelf has the same chance of being
infected.
Plus, in a grocery store, you have 1,000 other shoppers
going through and then the cashier has to, you know,
touch the goods that 1,000 other shoppers have
selected and touched.
So it's better just to reduce the number
of those interactions and just buy your groceries online
if you can.
But that's about it.
So you know, travel if necessary.
I think if you're armed with hand sanitizer and masks,
you can travel.
Just be careful, you know, with keeping your hands clean,
keeping your space clean.
And I would just end with this, that because 50% of people
with virus have no symptoms but will become immune
and because that means 95% of people
don't go to hospital, most people who
get the disease will recover at home in about a week.
The workforce is not threatened.
So, you know, there will be enough farmers and truck
drivers and store workers to distribute food and sell food.
So you don't have to go out and buy everything in sight.
So that's about it.
And in the next slide, as you guys have seen online,
there's a recipe for how to make your own hand
sanitizer if you're not able to buy any,
out of just ethanol or isopropanol and glycerol.
So that's it.
So I'm sorry to take so long.
Michael, this has been fantastic.
Thank you so much for taking time for us.
I wonder, do you have a couple of minutes for a few more
questions?
Sure.
OK, so we don't want to impose on your time.
But we do have a couple of questions.
Actually, we have a lot of questions.
Let me just start pick on a few.
Can you talk about the possibility of reinfection?
Yeah, you mean like getting an infected again
after you've cleared it.
Yes.
I think it's small.
You know, there was this worry that
was because a woman had recovered from the virus
and then left the hospital with a negative test
and then got worse and then got a positive test.
And it was taken as evidence that you could clear the virus
and get reinfected.
But it's probably just the negative test was negative.
I mean, she probably was fighting the virus pretty well,
but then the virus rebounded.
So it's more a case of just insufficient immune response
to the virus to begin with.
But if she cleared it the second time and it's really clear,
then it should be fine.
Your memory, you know, your B and T cells
take a week to ramp up.
But then your long-lived memory B cells take something
like a month to differentiate.
So it has to do with the switching of the types of IG,
immunoglobulin.
So there might be this window where the virus can come back.
But it's less likely that you're getting reinfected from outside
as it's just a low-level infection inside.
So along similar lines, what can you
share about the possibility of recovering this summer
and fall, all the students return to school,
and then we have a recurrence that's even worse than what
we have now?
Yeah, so I think one thing that public officials are not saying
yet-- well, they don't know what's going to happen--
is that these kinds of measures are
going to have to be kept in place
until a vaccine or a very, very effective treatment is made.
So you know, the RNA vaccine, so we'll probably
know if they're useful in six months.
But then, you know, it might take another few months to ramp
up production and distribution.
So we're not likely to be in time for a September
vaccination.
These drugs like chloroquine, remdesivir, you know,
chloroquine you don't want to really dose to everyone.
It's too toxic for that.
You might dose it to high-risk populations.
Remdesivir is likely to be in short supply.
It's probably only 50% effective anyway.
And so you don't want to rely on the drugs too much entirely.
[INAUDIBLE]
What are the barriers to manufacturing these drugs
faster and in larger quantities?
Chloroquine there's no barrier.
It's very simple.
The favipiravir also looks pretty simple.
I'm pretty sure they can be manufactured
for millions of people.
But the side effects are probably
such that you don't want to take it unless you really have to.
Yeah, and then remdesivir looks more complicated.
That's probably harder to synthesize.
But Gilead is upping the production.
But again, these drugs are not going
to cure everyone of disease.
And so they're more like hospitalization medicine.
So they only for the 5% of people go to the hospital.
And then maybe only 50% of those who take the drug
will still be able to fight it off.
And that's just a guess at the moment.
It may be 75%.
So it's still going to be a burden on the hospitals
unless we continue to be careful.
What I would like to see is us returning to work,
but just being more careful like people are in Asia.
I think it's a public education problem.
You know, in the US, the local authorities don't even
bother trying a course of education
before a course of lockdown.
Because they've seen how people don't
react to proclamations of staying away from crowded
places.
But, you know, in Asia you can do it.
People are maybe much more just much more scared of viruses,
more used to working in tight environments, used
to things like wearing a face mask and sanitizing spaces.
Can you also address briefly the testing bottleneck and what
we can do to test more?
I admire how Korea tested so many people.
And they have tests that come around in 50 minutes, I hear.
What can we do here?
Yeah, I don't know.
Yeah, I'm not sure about the 50 minute test.
I really don't know.
Maybe there is.
But that would be amazing.
What I saw in articles about Korea
is that they, first of all, the companies
took their own initiative to develop tests early.
And then the health ministry coordinated with many companies
to roll out multiple tests at once
because they knew it might be a big thing and they wanted
to be ready.
You know, Korea gets a lot of visitors from China.
So they probably suspected they would get a lot of infections
from China.
In the end, they got most of their infections
from one patient, patient 31, who infected people
through her church.
So they were ready for that.
You know, in the US, part of it is the federal nature
of the government.
Korea is still a relatively small country
of 40 million people.
OK, so what's funny is when China locked down
Wuhan in the province of Hubei, that was 50 million people.
And I just read an interview by Dr. Fauci where he was asked,
would that ever happen in the US?
He was like no, that's impossible.
We could never do that.
So now we've locked down California.
That's 40 million people.
It's the size of Korea.
So it's possible to do it.
But in the US we have 320 million
and the federal government.
And as we know, we have a federal government that's been,
one, starved of resources, and two,
ideologically reluctant to spend a dime on any public good.
So it's just a complete failure of our political system.
Well, I mean, let's be honest, it's
a complete failure of one set of people in charge.
Michael, you have a lot of captive computer scientists
and engineers on this call.
What can computing and engineering
do to help with this difficult moment in time?
Yeah, that's a good question.
Not being a computer scientist myself,
you know, I don't want to give advice.
I think you guys will figure it out.
You know, it's-- you know, making cheap ventilators is
certainly one possible solution.
You know, using computer science to work with biologists
to screen for drugs or analyze genomes
is another possible solution.
Yeah, I think talk to your colleagues.
Talk to your colleagues in biology.
It's harder to do that now that we
can't gather in the cafeteria.
But you know, we'll have a few weeks of being at home.
It's possible to urge your department chairs
to send messages to their other colleagues
in other departments asking if anyone
needs any assistance in their research efforts.
Michael, thank you so much.
If you get any computational challenges,
please reach out to us.
We would love to help.
And we really appreciate you sharing your time
and your knowledge with us.
Thank you.
I'm gonna clap on behalf of everyone.
Thank you very much.
Yeah, I appreciate your time.
Thank you so much.
OK, signing off.
Thank you.
Bye, everyone.
Thank you for joining us.
I hope you enjoyed the presentation and you learned--