On the 3.2kbytes in Influenza A

http://www.bunniestudios.com/blog/?p=353

by blasdel • 17 years ago

142 28 17 years ago

bunniestudios.com

ars3• 17 years ago

Since this virus mutates so often, shouldn't they sequence a few hundred versions of it?

And clearly I need to install a firewall to close some of those open ports.

robinhouston• 17 years ago

At the institute where I work, we're sequencing H1N1 isolates from (IIRC) fifty UK patients. There are also emergency plans to start sequencing flu samples on a massive scale if it turns into a major epidemic. I would be astonished if large sequencing centres in other countries were not doing similar things.

In other words: yes you're right, and it's already being done.

jballanc6• 17 years ago

http://www.ncbi.nlm.nih.gov/genomes/FLU/FLU.html

There are literally thousands of strains that have been sequenced.

silentOpen• 17 years ago

So, those of us with the mutation that causes lungs to ignore the H5 port would have a better chance of surviving an Avian flu infection, whereas as those of us that open port H5 on the lungs have no chance to survive make your time / all your base pairs are belong to H5N1.

jbyers• 17 years ago

That's the most interesting not-really-about-computers article I've read on HN in the last two years.

rw• 17 years ago

(awesome? bioinformatics)

kylec• 17 years ago

bmelton• 17 years ago

For your perusal: http://www.kk.org/thetechnium/archives/2009/04/inevitable_mi...

jhickner• 17 years ago

If you thought of organisms as computers with IP addresses, each functional group of cells in the organism would be listening to the environment through its own active port. So, as port 25 maps specifically to SMTP services on a computer, port H1 maps specifically to the windpipe region on a human. Interestingly, the same port H1 maps to the intestinal tract on a bird.

Could anyone expand a bit on this subject? I found this completely fascinating, but I'm having trouble finding further reading on the underlying mechanics of "ports", as the article calls them.

xel02• 17 years ago

Different species have a different type from the protein family Hemagglutinin(HA)(they all look similar to each other) these proteins are similar to services offered on a port. Viruses like Influenza target HA but only of certain types. In the case of Influenza HA doesn't actually act as a port, its not designed to allow things to go into the cell, the virus however uses HA as a recognition marker for where it will attack.

As a side note H1N1 means HA type 1, and Neuraminidase (NA) type 1. NA is a necessary protein for Influenza to leave the cell after replication.

jballanc6• 17 years ago

Not entirely correct. Hemagglutinin is the protein on the virus coat. It binds to sialic acid on the surface of the cells. Sialic acids can have different glycosidic linkages. In humans, the predominant linkage type is alpha-1,6. In birds the predominant linkage is alpha-1,4. Pigs have a fairly even mix of both alpha-1,4 and alpha-1,6 which explains why they make good hosts for recombination.

imajes2• 17 years ago

As i understand it, your body is made of various different types of cells. Now, a virus has to be able to mate with these cells to break through and then change the RNA of the cell so it can replicate- remember, viruses are sub-cell agents and need to cause mutations of good cells to replicate.

So given your body is made of different types of cells, viruses only have the keys to certain cells - where there's enough in common in their RNA to mate. This is kind of what's meant by the ports the author describes- H1N1 has enough similar encoding to the cells in your throat, nose and sinuses to mate and cause mutations.

ars3• 17 years ago

I don't believe this is correct. First a virus does not cause a cell to mutate. It's more akin to handing the cell a new set of work orders, but it doesn't mutate the cell's DNA.

Second, I think the "port" is not that the RNA is similar, but rather the virus has a protein key on it's surface that physically allows it to enter the cell.

In the micro world things move by atomic forces. In this case the virus has a protein on it's surface that is the same shape as a protein on the surface of the cell. The same shape means the electrical charges line up, and the virus is drawn toward the cell by electrostatic forces.

If the charge pattern does not match up, the virus will not be drawn toward the cell. Remember the virus has no ability to move.

jballanc6• 17 years ago

Well, this doesn't really get at the heart of the article (or the matter), but to be rather pedantic, the virus genome encodes much more than just 3.2KB. Keep in mind that most of what the genome accomplishes is done with proteins. In that case, you have to take codons (sets of 3 bases) and convert to one of the 20 amino acids. This effectively takes 64 bits of information and reduces it to 20...

EXCEPT that proteins are 3-dimensional entities. Their exact 3D configuration is encoded in the amino acid sequence encoded by the DNA. We still don't fully understand how this encoding works, but it is, in effect, the most awesome compression algorithm imaginable!

sketerpot• 17 years ago

There's some redundancy in the mapping of codons to amino acids. Some of the most common point mutations will cause a codon to map to the same amino acid, or a functionally similar one. That's some information that would be lost if you represented DNA as an amino acid sequence.

Robin_Message• 17 years ago

(Disclaimer: I am not a biologist)

DNA can only have four bases (ACGT), so each DNA base can only encode 2 bits of information. 20 amino acids could be encoded in 4 and a bit bits, not 20 bits (actually, you need a stop code, so 21 codes are needed).

As it is, DNA uses 3 bases, 6 bits to encode each amino acid, so there is a little redundancy. See http://en.wikipedia.org/wiki/Genetic_code#RNA_codon_table for the mapping.

Proteins are 3D entities, but their shape is determined by the amino acids, which are determined by the DNA - so there is no awesome compression here. To put it another way, the words "booktitle:Dune" encode a novel, which is much more than 14 bytes, but I cannot encode any arbitrary stream of bytes in this way.

jballanc6• 17 years ago

Please, please, PLEASE Don't take the author's advice of trying to catch H1N1!!! I realize that comment was tongue-in-cheek, but still... There are a lot of people who do believe that they'll be better off catching it, and this is only likely to make things worse.

imajes2• 17 years ago

well, since most of us get flu at least once every couple of years, it's kind of hard to avoid H1N1, the most popular form of influenza going around. Of course, if we were to catch the very specific form of H1N1 with the genetic coding that matches the swine flu epidemic, that'd be worse.

It's worth remembering that in almost all cases, it's not the flu that kills you - it's the onset pneumonia, diarrhea and dehydration that takes you down. The best survivalist method is to remind yourself how to treat cold/flu symptoms, stock up on meds, and sweat it out. Helps if your health insurance is up to date too. :)

jballanc6• 17 years ago

True, many will become infected naturally. The problem is that the antigenic shift of influenza is affected by transmission rate. That is, the more times a virus gets passed from person to person, the more it can mutate. The vaccines that are developed are based on a "normal" rate of antigenic shift. If large groups of people try to get infected, there is a risk that they will throw off this rate.

kqr2• 17 years ago

Interesting use of computer analogies:

  DNA		 : Program stored on disk
  RNA		 : Program loaded in RAM
  Amino Acid	 : Pixel in a frame buffer
  Protein	 : Image output from the RNA program
  Organism	 : Computer with an IP address
  Functional     : An application which listens to a particular socket
   Group of Cells

eru• 17 years ago

Yes interesting. Though Proteins are much more active than images. Perhaps an analogy with robots would be more apt.

stuntgoat2• 17 years ago

Joel, can I help you build a stackoverflow for virologists, please? I want to call it antigenoverflow.

flinc• 17 years ago

the best resource on H1N1 (and virology in general) I've found to counter the news hysteria http://www.twiv.tv/

On the 3.2kbytes in Influenza A

Related Stories