3 ms·

The J&J and AZ vaccines both use a similar viral vector with a modified gene. They target the same protein but just like with the mRNA vaccine they usually do

The J&J and AZ vaccines both use a similar viral vector with a modified gene. They target the same protein but just like with the mRNA vaccine they usually do not encode the entire protein but rather target specific amino acid chains in various regions of interest on the protein itself. Other differences in the vaccine would be the level and type of immune response generated in general which will be based on the specific viral vector used, dose and regiment, adjuvants etc. The study which measured the J&J efficacy also used a different methodology to this one. Overall this is a very limited study and should be taken in context. Another key part is that the T-cell response for the AZ vaccine group seems to remain intact. >Although the correlation between antibody response and vaccine efficacy is high, which suggests that the neutralizing antibody response is important, T-cell responses may contribute to protection from Covid-19 even in the presence of lower neutralizing antibody titers.32 In a post hoc analysis reported here, we found that in spike-specific T cells that expanded after vaccination with ChAdOx1 nCoV-19, the majority of antigens and epitopes remained intact in recognition of the B.1.351 variant. Until we have data on actual severe cases it seems that the AZ vaccine is still effective at preventing hospitalization and deaths. There were also no tracking of non-symptomatic cases in the study so the actual prevalence of infections in either group is unkown.

raducu 6y ago
Thanks, somehow with my software engineering mindset, I thought they somehow just copy&pasted the whole ARN sequence for encoding the spike protein from one virus to the other, and the end result should have been the same, but if its only parts of the S protein, then it makes sense that the outcome will be different.
- nicoburns 6y ago
  If you want a software analogy, then I think it's more like using regular expressions to try to match inconsistent data, where new data (which is potentially inconsistent in new and exciting ways) is coming in all the time.
  - dboreham 6y ago
    Like a virus scanner.
    - dogma1138 6y ago
      That’s a pretty good analogy the signature isn’t often not a simple hash of the malware but rather a more fuzzy signature based on core functions that are harder to alter without loss of function.
- dogma1138 6y ago
  I don’t know if they even have enough room to encode the entire protein, that said even if they do it’s quite likely better to focus on specific amino acid chains that relate to specific structural areas of the protein that serve a critical function like interfacing with the receptors these aren’t likely to mutate without loss of function and it increases the likelihood of more targeted immune response. If for example each vaccine selected 30 amino acid chains say 15 common ones and 15 that are varied between vaccines you can end up with vaccines that are more effective against certain mutations than others. Realistically COVID-19 isn’t going anywhere and it will likely become very much endemic and will require a yearly vaccine especially for risk groups just like influenza.
rossdavidh 6y ago
Well now that casts this in a significantly different light. I have read that T-cell response is much more labor intensive to measure than antibodies, but it seems to not uncommonly tell a very different story.
- nradov 6y ago
  That is starting to change. The FDA recently authorized Adaptive Biotechnology’s T cell assay. It's still more expensive than an antibody test. https://www.t-detect.com/ https://www.t-detect.com/

raducu 6y ago

Thanks, somehow with my software engineering mindset, I thought they somehow just copy&pasted the whole ARN sequence for encoding the spike protein from one virus to the other, and the end result should have been the same, but if its only parts of the S protein, then it makes sense that the outcome will be different.

nicoburns 6y ago
If you want a software analogy, then I think it's more like using regular expressions to try to match inconsistent data, where new data (which is potentially inconsistent in new and exciting ways) is coming in all the time.
- dboreham 6y ago
  Like a virus scanner.
  - dogma1138 6y ago
    That’s a pretty good analogy the signature isn’t often not a simple hash of the malware but rather a more fuzzy signature based on core functions that are harder to alter without loss of function.
dogma1138 6y ago
I don’t know if they even have enough room to encode the entire protein, that said even if they do it’s quite likely better to focus on specific amino acid chains that relate to specific structural areas of the protein that serve a critical function like interfacing with the receptors these aren’t likely to mutate without loss of function and it increases the likelihood of more targeted immune response. If for example each vaccine selected 30 amino acid chains say 15 common ones and 15 that are varied between vaccines you can end up with vaccines that are more effective against certain mutations than others. Realistically COVID-19 isn’t going anywhere and it will likely become very much endemic and will require a yearly vaccine especially for risk groups just like influenza.

rossdavidh 6y ago

Well now that casts this in a significantly different light. I have read that T-cell response is much more labor intensive to measure than antibodies, but it seems to not uncommonly tell a very different story.

nradov 6y ago
That is starting to change. The FDA recently authorized Adaptive Biotechnology’s T cell assay. It's still more expensive than an antibody test. https://www.t-detect.com/ https://www.t-detect.com/