HI tests and antigenic mapping.
While it seems such an obvious thing to do I was not, until recently, aware of antigenic mapping but it is a much more powerful tool for understanding vaccine scope than HI tables or cladograms. It also showed me something – at a glance – which I had not been fully aware of. The something being the discontinuous nature of strain drift in seasonal flus. Looking at sequences and generating phylogenic trees had left me with the mistaken ‘gut feeling’ that sequence drift was largely random and radial, in that it spread in all directions from the dominant wild strain with occasional changes of the meander's direction, rather Brownian motion like.
This map of H3N2 from its pandemic emergence in 1968 to 2002 shows a more complex picture with a definitely more ‘clumpy’ progression.
While this is new to me it appears to be something of genuine use IF we can get access to raw HI data in the same way as it is possible to get access to some of the sequence data via Genebank and GISAID. So my first question is ‘does anyone know of a repository of HI test data which is not behind a firewall?’ There is a website with more maps and it appears they will be releasing some analysis software soon.
This is an explanation of the grid scale from their site which has many other antigenic maps including H1N1(2009).
Had we been able to link the Lviv/N2 and N6 sequence data to their respective HI tests and this mapping technology a lot of the speculation in the low reactor thread would have been moot.
I would like to do a bit of linking and hat tipping. To gsgs for providing the link to this youtube interview between Vincent Racaniello and Fredrick Hayden which led me to Vincent’s interview with Derek Smith who is on the WHO vaccine selection committee and behind the antigenic mapping software. I strongly recommend watching this as it explains much about the selection procedure, viral evolution and mapping.
This link is to Vincent Racaniello’s excellent introduction to HI.
This to an old post of mine which I hope adds a little background on the path between sequence and immune response.
To a more recent post with some basic info on phylogenic trees and the importance of a temporal component and data linking to sequence data to which HI tests should be added.
This is to a longish video by Hans Rosling which has a great section on novel methods of data presentation and the future of realtime data analysis – all good stuff.
As always your take, thoughts, comments and links are all welcome.
JJ
While it seems such an obvious thing to do I was not, until recently, aware of antigenic mapping but it is a much more powerful tool for understanding vaccine scope than HI tables or cladograms. It also showed me something – at a glance – which I had not been fully aware of. The something being the discontinuous nature of strain drift in seasonal flus. Looking at sequences and generating phylogenic trees had left me with the mistaken ‘gut feeling’ that sequence drift was largely random and radial, in that it spread in all directions from the dominant wild strain with occasional changes of the meander's direction, rather Brownian motion like.
This map of H3N2 from its pandemic emergence in 1968 to 2002 shows a more complex picture with a definitely more ‘clumpy’ progression.
While this is new to me it appears to be something of genuine use IF we can get access to raw HI data in the same way as it is possible to get access to some of the sequence data via Genebank and GISAID. So my first question is ‘does anyone know of a repository of HI test data which is not behind a firewall?’ There is a website with more maps and it appears they will be releasing some analysis software soon.
This is an explanation of the grid scale from their site which has many other antigenic maps including H1N1(2009).
From the above it can be seen that a circle with a radius of two or three squares roughly equates to effective vaccine coverage and by selecting a virus from the near the centre of a coloured clump does, for most clumps, provide fairly good coverage. What is not show is the collection dates on the individual HI tests. The date of the representative sequence for each clump is given and these dates show clump-jumps every few years but it is not clear if any of the, for example, BE92 clump were actually collected later than 1995 and would be co-circulating with WU95 and therefore – assuming a vaccine change to WU95 like – would be outside the new vaccine scope.
Had we been able to link the Lviv/N2 and N6 sequence data to their respective HI tests and this mapping technology a lot of the speculation in the low reactor thread would have been moot.
I would like to do a bit of linking and hat tipping. To gsgs for providing the link to this youtube interview between Vincent Racaniello and Fredrick Hayden which led me to Vincent’s interview with Derek Smith who is on the WHO vaccine selection committee and behind the antigenic mapping software. I strongly recommend watching this as it explains much about the selection procedure, viral evolution and mapping.
This link is to Vincent Racaniello’s excellent introduction to HI.
This to an old post of mine which I hope adds a little background on the path between sequence and immune response.
To a more recent post with some basic info on phylogenic trees and the importance of a temporal component and data linking to sequence data to which HI tests should be added.
This is to a longish video by Hans Rosling which has a great section on novel methods of data presentation and the future of realtime data analysis – all good stuff.
As always your take, thoughts, comments and links are all welcome.
JJ