RByers Lab Benchtop view

Post Archive 2022

← All Years

Wow, I wonā€™t pretend to have the answers but itā€™s hard to believe that the questions being raised by @RADVACproject arenā€™t getting more serious consideration and widespread discussion. https://youtu.be/1Tzj76m1n1A

If you were the one standing at this switch, how quickly would you pull it? Iā€™m glad Iā€™m not standing there. But perhaps thatā€™s the problem - weā€™d all rather share responsibility for millions of preventable deaths than risk personal blame for a few.

12 years ago most experts felt it would be reckless to update major web browsers more often than 6-12 months. Now 4-6 weeks is the norm.
Itā€™s far from a perfect analogy, but perhaps thereā€™s something to learn from this for vaccines? @RADVACproject https://radvac.org/ 1/11

  1. Chrome led this ā€œevergreen browser revolutionā€ by taking the controversial position that it was more important to be able to respond to issues very quickly than to have maximum confidence in quality before releasing. A fast update culture can reduce net harm. 2/11
  1. Updates happen in a staged fashion with careful attention to usage and performance metrics. If thereā€™s an anomaly we canā€™t explain, updates are paused (or, worst case, recalled) while we figure it out. 3/11
  1. Lab testing is of course still critical. We do as much as we reasonably can WITHIN our release timeframe. 4/11
  1. Rapid and tight feedback loops are essential. If something goes wrong, users can file bugs with a system in place such that the right developer can hear about it, and potentially take immediate corrective action. 5/11
  1. The Chrome team created a culture and set of processes to ensure that emergencies would be dealt with effectively in a matter of hours. This includes ā€œblameless post-mortemsā€ to constantly get better at emergency response. 6/11
  1. Risky changes are done as random A/B trials starting with a small population and ramping up only when metrics look good. Yes, that means Chrome is always experimenting on average users and some users are harmed as a result! 7/11
  1. All users are given the choice to opt-in to even higher risk models (daily-update canary channel, weekly update dev channel, etc.) and feedback from those populations are essential to decision making for the larger user population. 8/11
  1. Finally, Chrome was released open source as the Chromium project so that when we inevitably make mistakes or have blind spots, others can step in and correct them by competition (forks) and/or cooperation (upstream contributions). Users can choose more stable browsers. 9/11

Ok, I know youā€™re outraged that Iā€™m suggesting maybe doing things with vaccines which would cause some harm. The important question is how much more or less harm than status quo?
Millions died while we waited for COVID vaccine trials! https://www.youtube.com/watch?v=1Tzj76m1n1A 10/11

Oh and if you think this fast-update model canā€™t possibly work for something with life-and-death consequences, then you probably donā€™t yet own a Tesla. Itā€™s just a matter of time before most cars will follow this ā€œevergreenā€ model too. 11/11

Note that something has gotten much worse in the past two weeks. Compare to hospitalization rates throughout the pandemic. 13% of all children <5 who have been hospitalized with COVID have been hospitalized in the last 2 weeks, compared to 3% for all age groups combined. Omicron?

Earlier this week I emailed a rapid testing expert to ask if they had data on throat vs. nasal swabbing. They told me no but they had symptoms themselves for the past two days but had tested negative in the nose. They did a throat swab and it was positive! Another nasal was neg.

I know itā€™s too small of a sample, but that makes two of two people Iā€™ve asked getting throat positive at least a day ahead of nasal. Its highly unlikely this is coincidence.

It feels like we, as a society, have been given 2 years to learn and practice infection control, and now weā€™re all being put to the test. Trial by fire! How are we doing?

Ottawa wastewater data is the best signal we have into COVID prevalence in Ontario. After being optimistic for a few days, Iā€™m now quite confident weā€™re not seeing the dramatic drop in prevalence many are claiming.
https://613covid.ca/wastewater/

Note that some provinces, at least Ontario, have also procured millions of tests on their own and so this is just a lower bound on the availability of government funded tests in each province.

Comparing Canada and USA for rapid test approval. What changed in the USA on Oct 21, 2021 and how can we effect a similar change in Canada?

USA Source: https://www.fda.gov/medical-devices/coronavirus-disease-2019-covid-19-emergency-use-authorizations-medical-devices/in-vitro-diagnostics-euas-antigen-diagnostic-tests-sars-cov-2
Canada Source: https://www.canada.ca/en/health-canada/services/drugs-health-products/covid19-industry/medical-devices/authorized/list.html https://x.com/RickByersLab/status/1484634454568189964

Since you all seem to like when I post graphs of interesting data, hereā€™s another one for you.

In the future, when we have outbreaks vaccines will be mailed to every house (eg. in nasal form). When new variants arise, we will test and update our vaccine formulations within weeks. Wastewater testing will mean the right vaccines will be sent to the right places quickly.

Most people will take these vaccines and rarely feel the need to alter their behavior otherwise. Weā€™ll look back on our failure to handle pandemics like COVID-19 like we now look back on sailors who struggled with scurvy due to lacking simple vitamin C supplements.

This may sound like sci-fi, but itā€™s not really. We probably have the capacity to build up the technology today. So much physical and economic suffering due mainly to a lack of vision and willpower. šŸ˜¢
https://RadVac.org @RADVACproject

True, but it took over 50 years from the invention of seatbelts for them to be mandated in Ontario (8 years after the USA). Still many believed they were safer without them. Apparently we are terrible at adopting life saving tools when they involve the slightest inconvenience! https://x.com/DrShaneRRR/status/1489339895768555523

When comparing test sensitivities itā€™s critical to use a calibrated viral load cutoff like 10^6 RNA copies/ml. Ct values themselves are almost meaningless. Hereā€™s why with an example:
1/8

I was researching the sensitivity of the Innova rapid antigen tests used widely across the UK. Low viral load samples need to be ignored because they generally donā€™t represent contagious levels of virus. Many studies use a PCR cutoff around Ct<25.
2/8

No! Ct values are relative to a particular device and test protocol (eg. A sample diluted 2x more will have a Ct one higher). Sadly most studies donā€™t publish their Ct calibration in absolute terms like RNA copies per ml. But these two do!
5/8

German study: Ct=25 is 10^6 RNA copies/ml.
Liverpool study: Ct=24.4 is 10^4 RNA copies/ml.
Thatā€™s a 100x fold difference!
The Liverpool study uses Ct=18.3 for 10^6 RNA copies/ml.
6/8

And sure enough at Ct<18.3 the Liverpool study shows a sensitivity of 91%, exactly consistent with the German study!
7/8

So in conclusion, donā€™t trust any test sensitivity number unless itā€™s accompanied by a cutoff in terms of RNA copy number.

>10^6 RNA copies/ml is a good standard threshold for contagious samples.
8/8

The Cue is a molecular test similar to lab-based PCR tests and the Lucira check-it. Theyā€™re more sensitive than antigen tests, but at >$60 per test, for screening purposes youā€™re probably better off doing antigen tests 3x as often for half the price. Still, itā€™s interesting.

There is a nifty app that controls the device and walks you through the process.

It looks like thereā€™s only a single reaction chamber with some sort of electrical sensor that sits above the heater in the reader. Fluid flows from two chambers, through the swab and into a sponge.

Iā€™m guessing itā€™s also RT-LAMP but with some sort of electrical measure of the pH change. But it claims to also test for human RNaseP, so a swab without a sample will return invalid. I have no idea how theyā€™re obtaining multiplexing in one reaction well with no optics. Thoughts?

I managed to pry it apart further. The chip looks to be a 24C04, 4kbit EEPROM worth about $0.32 USD. So probably just for storing the cartridge serial number, maybe assay configuration parameters so the device can work offline with new tests.

My wife was asked at the grocery store checkout if sheā€™d like some free rapid tests and the clerk was clearly disappointed when she said ā€œno thanksā€. šŸ™‚
My wife explained that we have purchased lots and want to make sure thereā€™s enough tax-funded tests for those less privileged.

According to the clerk, most people are glad to take them and the store has plenty of supply available. Can it be that weā€™ve nearly reached the elusive goal of universal access to rapid tests in Ontario? šŸŽ‰ @C19TestFinders

In this pre-omicron program it would have cost about $4,000 in tests to catch one case and avoid the downstream transmissions. How should we, as a society, decide what the threshold for a good ROI is?

I got my 12yo boosted today despite being a little shy of the 6 month interval. The clinic didnā€™t even mention the 6 month interval.
With kids back at school without mask mandates next week, the risk tradeoff seems pretty damn clear to me. Iā€™m glad I didnā€™t have to fight for it.

My sonā€™s report from the first day of grade 10 without mask mandates:
ā€œIn business class 20% of the kids were wearing masks. In pre-AP math class it was 80%ā€
šŸ¤”

Also there was a significant decrease in mask wearing throughout the day, with math class being a huge outlier at the end of the day.

A COVID transmission anecdote and live test of my familyā€™s airborne transmission precautions:

My sonā€™s teenage friend rides in a car for ~30 min, masked the whole time. The next day the driver tests positive. Teen is triple vaxxed.

At day two and three post exposure the teen has a negative rapid antigen test. They had plans to come to our house Sunday (with everyone present doing rapid tests) but given the exposure I ask to postpone. Rapid tests arenā€™t perfect!

Around four days post exposure the teen has a negative PCR test. Thatā€™s gotta mean the risk is low, right?

Being masked, the exposure doesnā€™t count as a ā€œclose contactā€ so the teen is still at school with my son.

Seven days post exposure the teen does another rapid antigen test and this time itā€™s positive! No symptoms yet.

My son has been sitting next to the teen and eating lunch with them for the past three days. Whatā€™s the probability my son is now infected?

Well, my son has made some key decisions (his, not mine):

  • Got boosted the first day he was eligible, 6 weeks ago
  • Always wears an N95 mask at school
  • Eats lunch outside, keeping >1m distant from others (even in our record cold temps, which my wife and I said was nuts).

I figure heā€™s got a pretty good shot at having protected himself, but with such prolonged exposure the risk has still got to be significant.

My son will be doing rapid tests daily to help reduce the risk of him spreading it to someone else at school. The first is negative.

My son had another negative rapid antigen test and this morning a negative rapid molecular test (Lucira check-it).

For the record, my son continued to test negative on daily RATs. We stopped regular testing about 4 days after end of symptoms (though I admit to testing only every other day during those 4 days). We also just did an antibody test and saw no hint of IgM (only IgG from vaccine).

I suspect most everyone was afraid of policy and PR implications of selling SARS-CoV-2 PCR kits to the public. That @miniPCR is now OK doing this signals to me a real opening up of legitimate citizen science. Next up: all respiratory viruses?
Srsly, what are we afraid of? šŸ™‚ 2/2

To be fair, companies like @biomeme and @chaibio were already ahead of the game at the start of the pandemic. But selling devices and kits targetted at (and priced for) the highschool classroom is a step function increase in scale and acceptance.

UPDATE: Looks like I spoke too soon. Digging deeper into the instructions I see this isnā€™t a real virus test kit (no RNA extraction, no reverse-transcriptase). It comes with DNA samples, presumably synthetic. Thatā€™s a shame šŸ˜¢.

Weā€™re already experimenting with COVID transmission in pretty much every single classroom in North America and Europe. Would it really be any more risk to at least get some science education out of it? šŸ˜

These graphs did an excellent job explaining the value of vaccination to Ontarians right up until December. Now theyā€™re really failing to capture the most valuable opportunity. What gives?

I wonder how many other pandemics weā€™ve narrowly avoided by one of the first transmissions being interrupted somehow?

What small thing would it have taken to prevent these last two and a half years of hell?

My daughter got some new pets today: genetically engineered Tetra GloFishĀ®ļø
https://en.wikipedia.org/wiki/GloFish

What excites me is not that these exist (standard lab work these days), but that PetSmart has a whole wall of them and people are by-and-large cool with it! šŸ¤“

This certainly beats the home CRISPR kit we were planning on trying on bacteria: https://www.the-odin.com/diy-crispr-kit/.

Iā€™m picturing what PetSmart might look like in 20 years. Are we at the dawn of a modern ā€œCambrian Explosionā€ in biodiversity? Yes, gotta be mindful of ethics and safety.

At this point, isnā€™t this the MOST important indicator for the impact of the pandemic? Why arenā€™t more people talking about it and pushing to figure out how to get us back to >95%? @imgrund @DFisman

Prediction: 5 years from now when the impact is undeniably obvious, everyone will complain: why didnā€™t healthcare experts warn us of the risks of long COVID and explain how a simple N95 could massively reduce our risk without lockdowns?

I couldnā€™t keep asking my son to miss life experiences, so I took him to paleontology camp in the US.
Pro: 3 awesome days of learning from paleontologists in the field.
Con: COVID+ on day 4, now stuck in a hotel in the middle of nowhere for at least 10 days šŸ˜„.

Itā€™s crazy that antigen tests are available in most pharmacies in the US yet only in one or two packs!

I order boxes of 25 at home, and Ontario gives out 5 packs in grocery stores. Buying in the US feels like such a waste of packaging, not to mention higher prices!

Iā€™m sad that when COVID finally caught up with me, I was away from my lab and so couldnā€™t collect daily samples for PCR Ct monitoring. I did collect one sample near the end, Iā€™m hoping to figure out how to get it sequenced so I can see where we fell on https://nextstrain.org/ncov/gisaid/global/6m.

Finally got around to running qPCR on my personal COVID sample taken on my last day of antigen positivity.
Ct of 29 on CDC N1 and N2 targets (green lines, red is internal control, 3rd sample is negative control).

Next step: figure out sequencing and variant analysis!

Only 6% of cases were asymptomatic and only 25% shed culturable virus pre-symptomatically. This is a pre-omicron largely unvaccinated population, but from earlier studies Iā€™d expect ~50%. 2/6

As expected, RAT sensitivity correlated with viral load. However this correlation disappeared in the period after peak viral RNA! 3/6

There were a few more instances of culturable virus with negative RAT than I would have hoped or predicted. The paper calls these ā€œinfectiousā€, but as @profvrr likes to remind us: ā€œculturability doesnā€™t necessarily imply infectiousnessā€. 4/6

There were more cases of positive RAT days after virus was no longer culturable than I would have expected. 5/6

We all know RATs are imperfect and itā€™s informative to see some better details quantifying that imperfection. But this study largely confirms the argument of @michaelmina_lab and others that frequent rapid antigen testing is the best practical way to predict contagiousness. 6/6

For my birthday my kids got me some supplies for my molecular biology lab. Iā€™m not sure if I should be offended or proud šŸ˜›.

In case itā€™s not obvious, for the record the shirt is a joke and I do indeed take lab safety seriously.

Had a dozen friends over, CO2 really spiked! This is in a large house with an air exchanger on medium, measuring with @airthings in an unused room.

Itā€™s a good reminder of how much weā€™re breathing each otherā€™s air even when 6 feet apart! In this case, everyone did COVID RATs.

Got a new piece of lab equipment, a magnetic separation rack for DNA purification! šŸ˜
But something seems messed up with the market when the cheapest version, some plastic and three rare earth magnets, costs $300! šŸ˜‚
https://www.neb.ca/detail.php?id=7017

Looks like I should have searched Etsy or just made it myself! šŸ¤£

My intimate foe.
For 18 days you consumed me,
Now Iā€™ve laid you bare.

I attended a week-long conference with ~400 people. Every single one of us did a rapid test every morning and wore a mask when indoors (almost all N95). Many of us ate only outdoors. It really wasnā€™t hard at all and worth it to see people IRL! https://www.w3.org/2022/09/TPAC/health.html

I, and many others, would have still attended with fewer rules. Iā€™m sure many felt it was overkill for them personally. But we all knew someone at higher risk and so we were happy to take extra precautions to be as inclusive as possible.

Also the organizers did an awesome job ensuring people could participate in every meeting remotely. Being in person was better, but being engaged was possible regardless of willingness to travel.

The future is finally here!

In the past week Iā€™ve gotten two technologies Iā€™ve dreamed of owning for a decade: a self-driving car and a genome sequencer.

Neither is actually practical for widespread consumer usage yet. But the remaining journey now seems imminent and inevitable rather than being science-fiction.

Yesterday I hit my pre-decided conditions for ending my strict mask wearing at work.
Today someone I spent a bunch of time in a small meeting room with told me they tested strongly positive.šŸ¤¦ā€ā™‚ļø

But what is the guideline on when to START testing after exposure? After infection, the virus ā€œlatentā€ (replicating but not shedding) for the first 2-3 days, maybe up to a week or more. Testing on only day 0, 2 and 4 after exposure would give a dangerous false sense of security!

But it treats ā€œis currently infectedā€ as the only data point of value, perpetuating the terrible idea that home use tests are for ā€œdetecting an infectionā€ rather than the public health goal of minimizing the risk of onward transmission. In contrast see: https://x.com/RickByersLab/status/1561470940109733890.

What does the FDA even mean when they say ā€œinfectedā€? Eg. they say ā€œMolecular COVID-19 tests are generally expected to detect the SARS-CoV-2 virus at least 95% of the time when someone is infectedā€. But with a latent period of >2 days, this is likely mathematically impossible!

I think maybe when the FDA says ā€œinfectedā€ they mean ā€œshedding RNA as detectable by PCRā€, a circular definition! This is what has led to all the confusing and flawed reasoning evaluating antigen tests that @michaelmina_lab has been fighting against for years.

The overall message still reads like antigen tests have poor performance (especially for asymptomatic people) while lab PCR tests are perfect. Why doesnā€™t anyone talk about all the false negatives that occur in lab PCR tests performed during the unpredictable latent period?

The message should be: ā€œSince there is variability in time from exposure to becoming contagious and since no test is perfect, repeat testing over days 3-7 after exposure is necessary in order to conclude an individual is not likely to become contagious.ā€

And also ā€œTesting after the onset of symptoms or with a molecular test (instead of an antigen test) can help reduce this window of uncertainty by a day or two, but not eliminate it.ā€ [10/10]

My first run on a minION sequencer. If all goes to plan, later today Iā€™ll have the sequence of the SARS-CoV-2 virus that infected me!

24 million bases called in less than 30 min, should be more than enough for a 30kb genome!

Correction: using a more recent Nextstrain it more precisely classified my virus as BF.28, aka. BA.5.2.1.28.

Wore my @VaccinatedUS shirt to get my @pfizer bivalent BA.4/5 booster this morning. It turned out it was a new @ROWPublicHealth clinic and I was their very first customer! šŸŽ‰
The doctor was surprised that a software engineer had been able to sequence their own SARS-CoV-2 genomešŸ¤“

I asked @midjourney_ai to imagine me working in my lab. Here are some of my favourites. šŸ¤Æ

Not exactly accurate šŸ˜‚. But just imagine the power of AI-generated imagery for science communication if the accuracy continues to improve at itā€™s current pace and tools build the ability to refine specific details.

My son had a bad cold, but COVID tests were negative so we wondered what it was.šŸ¤’šŸ¤·
I just sequenced a piece of it and found it was Rhinovirus C, the ā€œNAT001ā€ strain!šŸ§¬
Someday this will be trivial and cheap to do at home. Knowledge is power!šŸ§ āš•ļø