Sunday, April 26, 2009

Swine Flu: Info-epidemiology

The recent outbreak of swine flu in Mexico which has now spread to other places (including my fair city) has spawned a plethora of websites and Google maps to try to help people track its spread. Ben Parr from Mashable has made a nice summary of some of the key ones, but my favorite so far has been Rod Page's Timemap, which allows users to see the spread through time on a map of the world (also see Page's post on this effort at iPhylo). Having been involved in a project myself from 2005 to 2007, which sought to merge viral genomics and GIS, it is really cool to see that some of the databases and information sources have finally come together in ways that are actually allowing for rapid dissemination of these types of data. Hopefully this virus can be contained very quickly - and hopefully this new field of info-epidemiology will help with that.


sergios-orestis kolokotronis said...

Maybe a reassorting strain? The CDC's journal Morbidity and Mortality Weekly Report (MMWR) reported on April 24 on the two SoCal cases that they are indeed swine influenzavirus A (H1N1). It is notable that these two cases appear to be epidemiologically unlinked yet genetically similar. Their "hemagglutinin (HA) gene sequences are similar to those of swine influenzaviruses that have circulated among U.S. pigs since ~1999; however, two genes coding for the neuraminidase (NA) and matrix (M) proteins are similar to corresponding genes of swine influenzaviruses of the Eurasian lineage" [Ref]. It's possible this is evidence of antigenic drift (video). Although since there is no formal national surveillance program for swine influenzaviruses to inform us on which strains are actually present in the North American pig population.

I don't know if I am missing this outbreak's sequences on public databases, but it'd be good for public health if independent groups of scientists could analyze sequences as they get produced to estimate mutation rates, the evolution of resistance/susceptibility to antiviral drugs, etc.

There is -yet- no apparent link between humans and contact with pigs, so maybe the sequences can help tease apart patterns.

Susan Perkins said...

Get on it, Sergios!

sergios-orestis kolokotronis said...

Correction 1: I meant that this could potentially be a case of intra-subtype reassortment – a combination of U.S. H1 and Eurasian N1. This is only what I consider to be a logical interpretation of the CDC's MMWR article (see above) on the 2 SoCal cases. I have no molecular subtyping information on any other case, therefore it would be irresponsible to generalize.

Correction 2: I meant to write "antigenic shift", not "drift".

Sorry about that. I can't seem to edit my own comment on Blogger. Let's hope the data becomes available soon.

Susan Perkins said...

Wow! This swine flu outbreak has the bloggers going crazy. Just received an alert from Genome web, which pointed me to this blog by Sandra Porter. Porter apparently downloaded some sequences off GenBank and used their tree-building methods (NJ only of course) to come to the conclusion that the new strains are related to those from an outbreak at a county fair in Ohio in 2007. Many people commented on that post, pointing out the multitude of problems with this analysis and cautioning that this relationship is not upheld so rushing off to publicize this is not a good idea. The moral of the story is probably: let trained systematists make trees. Just because a web application will let you do it doesn't mean you should.

Glor said...

I think this other blogger had a good idea, but she definitely should have been a little more circumspect before publicizing her results. Even if the neighbor-joining tree can be trusted it seems clear that sampling is an issue with her trees. I assumed there would be people doing this job with the appropriate methods, seems like a great opportunity for someone with training in phylogenetics to make a difference.

Susan Perkins said...

I might even happen to know someone...someone close...who is working on this very problem as we speak.

Anonymous said...

The importance of vaccine and preparedness has become more obvious in wake of the "swine flu virus". The CDC has developed a healthcare hand book to help combat a variety of Vaccine Preventable Diseases.

The CDC’s Epidemiology and Prevention of Vaccine-Preventable Diseases, 11th Edition (The Pink Book) – Just Released!

The new, 11th edition of the Centers for Disease Control and Prevention’s (CDC) Epidemiology and Prevention of Vaccine Preventable Diseases (The Pink Book) is now available from the Public Health Foundation (PHF). “The Pink Book” provides physicians, nurses, nurse practitioners, physician assistants, pharmacists, and others with comprehensive information on vaccine-preventable diseases. The new 11th edition contains the latest information and updates on immunization, including:

Revised principles of vaccination
Updated recommendations on immunization
New immunization strategies for healthcare practices and providers
Guidelines on vaccine safety
This essential resource is now available for ordering online by visiting the PHF online store at Ordering via mail, phone, fax, and purchase order is also available by calling PHF toll-free at (877)252-1200 for full instructions.

Dave said...


As a first pass, I tried doing a similar analysis, using the first 100 BLAST hits for each gene and a ML methodology. I'd definitely appreciate any feedback, especially from you guys.

For a second attempt I'm going to try rooting the tree and will also add the genes that
Taubenberger et al.
used to characterize the 1918 influenza virus. By the way, I think it's ironic that Taubenberger et al were arguing that the 1918 virus came from avians, when their chart shows that it unambiguously is close to classical swine flu.

I agree though, I think this is best left up to the experts -- I am mostly doing this for fun.

Specifically, Steve Salzberg et al. seem to know what they're doing.


Susan Perkins said...

A quick link to the great website of phylogenetic, phylogeographic, and demographic analyses of H1N1 by Holmes, Pybus, Rambaut et al.

iayork said...

I think it's ironic that Taubenberger et al were arguing that the 1918 virus came from avians, when their chart shows that it unambiguously is close to classical swine flu.You've got it backwards. The 1918 flu didn't come from swine flu. Swine flu came from the 1918 flu. That's why it looks like classic swine flu -- because it's the parent of swine flu. The order is avian -> 1918 human -> swine.

This has been known since the 1930s.

Susan Perkins said...

It clearly is a more complicated history that that. A new analysis out today, provides a hypothesis of how reassortment has likely produced the new H1N1 "swine flu" viruses we're seeing. A nice schematic is available here.

Natural Remedies said...

Good research Susan, Thanks for link.

Good Health Blog said...

thanks for the information..really gr8 help for not-so-tech savy bloggers :)..keep us updated.

Sujan Patricia said...

People with regular exposure to pigs are at increased risk of swine flu. More than 1100 people worldwide have died from swine flu since it emerged in Mexico and the US in April, according to the latest figures from the World.

Digital Scales said...

I can't fall into the entire "swine flu" craze because I think it's a wee melodramatic. Most people that come in contact with it have zero symptoms, or just sleep it off. Why we Americans have lost our minds over the swine flu is beyond me. How many deaths in the U.S. have been attributed to it??