Subscribe to TWiV

...with iTunes:

iTunes

...with Miro

Miro Video Player

...with web-based podcatchers:

add2netvibes

addtomyyahoo4

...with something else:

feed-icon-12x12-orange View RSS Feed

mail-icon-16x16 Email

Get more info on other podcatchers:

badge_juice

Letters

TWiV 90 Letters

Eric writes:

Hi Vincent,

After 86 episodes of TWiV, I am still loving every minute of the podcast and I am constantly impressed with how much I learn. For example, I have been mentoring an undergraduate student in the lab and we have had a rough two week stretch where none of her experiments have worked. I was racking my brain, trying to come up with some things that would help boost her confidence. On my way in to the lab on Thursday morning, I listened to you and Rich and Dr. Kiki; and Rich described how a good mentor guides the student in a direction that will allow them to succeed. I took that approach with her that morning and we had a great discussion about the project, what additional experiments she could try, and I was able to assure her that it was not her skills that were lacking. On Friday, she had two great successes that really boosted her confidence, and I felt like a good mentor! Thank you TWiV!

Tonight, I listened to Dark Matter with Eric Delwart, and I loved it. I have been experiencing many of the same things that Dr. Delwart described, and in fact I have 100s of hours of eyeball time invested in the bat virome project.

Thanks for your continued work with the podcast. I find it extremely helpful to hear about different viruses that I would normally have no exposure to.

Thanks,

Eric

David writes:

TWiV gang,

Several emails you’ve received have addressed the question, can computer viruses mutate? My take on this question is a qualified yes, but in a different way than biological viruses.

If I remember correctly, one of the emails you received answered this question as no, because randomly changing bits of a computer program almost invariably result in a broken program. I can vouch that this is true – I work in a group that designs computer chips, and any single bit change will cause the mutated program to go out into the weeds. However, mutations in biological systems, and as well as those that computer viruses undergo, happen at a different level.

By my way of thinking, changing a single bit in a computer program would be like changing a single atom in a biological compound. If you replace a Carbon atom with a Gold atom in Adenine, what you’ll probably end up with is an expensive non-functional molecule. Instead, mutations in DNA and RNA seem to happen at the molecular level, for example where an Adenine gets replaced by an Guanine, in the case of single nucleotide polymorphisms. Other mutations, such as copy number variations, occur at the level of a large collection of molecules. The point is that biological mutations occur at a higher level of organization than “bits”. This greatly increases the probability that the resulting mutants are viable (and sometimes “fitter”) than the original.

For computer viruses, I’m unaware of any comparable higher level of mutation inherent in computers. I’m not exactly how accurate computer based copying is, but I’ll hazard a guess that it is more accurate than biological copying. Errors in computer copying that I’ve typically encountered have been due to equipment failure (queue Drobo promotion here :) ). Because of this, when computer viruses change, it is because the computer virus itself purposely inserts variation into its offspring as they are created.

One way antivirus programs detect viruses is to look for virus signatures, sequences of bits in viruses that are characteristic of a virus or family of viruses. To evade detection, one tactic newer viruses use is to modify (“mutate”) their children to be unrecognizable to the antivirus software. The viral code is modified at a higher level of organization than bits, such as reordering instructions (analogous to SNPs, perhaps). In this way, some computer viruses do change (mutate if you will), but they self-mutate, rather than let nature do the mutating. For some more background, see http://en.wikipedia.org/wiki/Computer_virus#Stealth.

The Tierra program mentioned in one episode, mutates at both at the bit level and at a higher level of organization (swapping segments of code). The Tierra virtual environment does the mutating in this case.

This does beg one question – are there any biological entities that you are aware of that actively self-mutate?

Keep up the good work. I listen to several podcasts, and TWiV (and TWiP) are among the few that I make sure I have time and concentration to really listen to. I get a lot out of these podcasts.

– David

P.S. I guess the moral of the story is don’t mutate the bits, mutate the organization :)

 

Comments (0)

Collections (0)

American Society for Microbiology
2012 1752 N Street, N.W. • Washington, DC 20036-2904 • (202) 737-3600
American Society For Microbiology © 2014   |   Privacy Policy   |   Terms of Use