The other two have lots of fancy things done to the mRNA itself, including pseudouridine substitutions, 3 and 5 prime untranslated regions, and a poly-a tail. The payload comes from the NIH, and is modified by them as well with proline insertions to stabilize the spike in the prefusion form. All current EUA vaccines in the US (BioNTech, Moderna, Jansen), plus NovaVax (code not delivered to the user in that case) use some form of the NIH payload, as far as I can tell.

Edit: I forgot about the proprietary mRNA delivery systems as well, I don't really know much about those.

Interesting list, thanks! Do you know of any popular science description of these? I did biology at middle school and not since (but I continue to find the mechanics of low-level biology fascinating).

Should be a good start.

https://berthub.eu/articles/posts/part-2-reverse-engineering...

https://berthub.eu/articles/posts/reverse-engineering-source...

https://berthub.eu/articles/posts/curevac-vaccine-and-wonder...

https://berthub.eu/articles/posts/genetic-code-of-covid-19-v...

Thanks, these are really good reads.

I found this being a very good explanation for the technically inclined: https://berthub.eu/articles/posts/reverse-engineering-source...

Another good and accessible discussion of what went wrong with this vax is Derek Lowe's "In the Pipeline" blog:

https://blogs.sciencemag.org/pipeline/archives/2021/06/17/cu...

> I continue to be amazed at how small the sample sizes are

”…an interim analysis based on 134 COVID-19 cases in the study with about 40,000 volunteers in Europe and Latin America.”

The sample size is 40000 is it not? That’s as big as they come.

40,000 yes, big. But when looking at who got Covid, the numbers are tiny. In the end you are comparing who had vaccine/placebo on a group of O(100) people.

It seems these trials aren’t actively monitoring the participants to see who gets even mild Covid, they just look at who ends up sick and tests positive.

These Phase III vaccine trials also do weekly symptom monitoring. In order to qualify for lab evaluation for COVID-19, there's criteria for how many symptoms are required and for how long.

The Pfizer trial results were based on 170 confirmed cases but there were “3410 total cases of suspected, but unconfirmed covid-19 in the overall study population, 1594 occurred in the vaccine group vs. 1816 in the placebo group.”[1]

[1]: https://blogs.bmj.com/bmj/2021/01/04/peter-doshi-pfizer-and-...

> there's criteria for how many symptoms are required and for how long.

Interesting, why? In the UK at least, they now hand out Covid tests like candy.

> But when looking at who got Covid, the numbers are tiny.

True, of a sample size of 40,000 a small number got COVID. Thankfully it is big enough as the efficacy numbers have proven very accurate in real world measurements.

> Thankfully it is big enough as the efficacy numbers have proven very accurate in real world measurements.

But are you considering the fact that breakthrough infections are counted using comparitively low CT values in RT-PCR tests?

Do we know how it will affect the positive rates of the test?

> In the end you are comparing who had vaccine/placebo on a group of O(100) people.

I guess this is what people were saying about Absolute vs Relative risk reduction. If that is so, your explanation makes it much more clear.

You can presume that a large percentage of that 40,000 weren't exposed to covid19 during the study.

I remember a year ago the Chinese shutting down a covid19 study in China because their infection rates went to zero.

Could be interesting to see comparative studies of vaccines.

Right now the efficacy numbers seem be affected by when and where the studies were done.

There were ideas thrown around to do one big study of various vaccines instead of multiple ones. Didn't happen, probably was too much of an organizational challenge and would've required all companies to comply to a protocol made by a third party.

Scientifically it would've been the best idea.

The challenge now is that the more the world gets vaccinated the harder it gets to do any meaningful trial.

Why would the companies producing the vaccine have a say in this? Sure Pfizer might go “nah - we are happy with our trial - we don’t need another”. But come on ..,

> Why would the companies producing the vaccine have a say in this..

Because, if I am not mistaken, these studies are funded by the companies themselves.

Sure but they do them because they are required by law to do so.

Anyway. The public is paying for this in the end. If there is genuine suspicion that these efficacy numbers no longer hold, we should just sponsor a test of it. Otherwise we base our decisions on what to buy and how to roll it out on wrong data.

Sure, that is what I meant, they are required by law to show the benefit of the drug once.

So its really upto them if they want to do another trial, right? At least as per current law..

For all we know it's not actually that much worse than the vaccines currently in use since those were tested before the modern variants were in circulation. (The figures for the effectiveness of the other vaccines aren't entirely comparable - they're not from controlled studies and I don't think they cover one of the latest variants that made up a reasonable proportion of infections in this trial.) I wouldn't be surprised if it's still better than the non-mRNA Johnson & Johnson vaccine too...