Monday, October 11, 2004

Systems biology

It's the new new thing.

It's news in Wired

There's an Institute for Systems Biology

Nature Biotechnology features Systems Biology in a recent issue. The article on commercialization links to players. See the figures from the original article for details (subscription required).

Public companies with research programs on systems

Eli Lilly
Johnson & Johnson
Novartis
Novo Nordisk

Systems Biology Tool Companies

Accelrys
Genego
Genpathway
IBM
Ingenuity
Prosanos
TargetDiscovery

Private Systems Biology companies

beyondgenomics
entelos
bioseekinc
gene-networks
gnsbiotech
genomatica
genstruct
metabolic-explorer
optimata

The microbiome

Greetings "highly complex conglomerations of human, fungal, bacterial and viral cells"!

The concept of the superorganism is not new. The most familiar, and perhaps comfortable examples from classical biology are social insects like the honey bee and colonial protozoa like volvox. In both these cases, the superorganism is composed of multiple individuals of the same species, although for bees that individual is itself multicellular, while for volvox the individual corresponds to a single cell. The understanding that both as individuals and as societies we, as humans both constitute and are members of superorganisms is hardly new.


Consider another entry in the superorganism bestiary. There has been quiet discussion and acceptance of the fact that all mammals possess a microbiome, a nice word referring to the more than a thousand species of gut microorganisms whose population outnumbers the cells in the mammal harboring them. The number of genes in the microbiome is estimated in the millions, again dwarfing the 30,000 or so in a mammalian genome. The microbiome has been somewhat ignored, but indications are that is changing.

Articles in Science reporting genomic sequence of a symbiote and a pathogen from the human microbiome are summarized here.

An article in Nature Biotechnology introduces the basics of the microbiome in the context of predicting drug metabolism and toxicity.


Diet clearly has a major influence on many diseases and modulates the complex internal community of gut microorganisms. The particular microbial community in an individual mammalian host is referred to as the microbiome. These microorganisms, weighing up to 1 kg in a normal adult human, may total 100 trillion cells. This means that the 1,000-plus known species of symbionts probably contain more than 100 times as many genes as exist in the host. Together these interacting genomes can be considered to operate as a super-organism, with extensive coordination of metabolic and physiological responses, particularly at the gut-liver and the gut-immune system levels. Not all of these interactions are necessarily obligate but the degree of true physiological association is difficult to study as all mammals possess a microbiome. Indeed, because of the level of comanagement of many biological processes by alien symbiotic genomes, Xu et al. recently stated that "sequencing the components of the microbiome can be viewed as a logical albeit ambitious expansion of the human genome project"!


In Wired News this was reported somewhat sensationally as People Are Human-Bacteria Hybrid, rather than emphasizing the superorganism status of the microbiome per se. In any case, wider recognition of the true nature of our relationship (intimate and continuous) with the microbiome should result in expanding knowledge in this area, with corresponding benefit in terms of human health and comfort.

Another article describes greatly increased numbers of environmental sensor proteins, both secreted (sigma factors) as well as membrane spanning (histidine kinase linked receptors). I am unable to link directly to the article Xu et al, Trends in Microbiology 12,21 2004. Here is a web page that should be close by.

Technology making a difference

A Macarthur fellowship to MIT instructor Amy Smith, who provides engineering solutions to third world problems that are cheap, simple, maintainable and effective. All the trees are being cut down in Haiti to cook food? Provide a way to to turn abundant agricultural waste into charcoal. Is it taking all day to grind grain for daily bread? Provide a cheaper non motor driven mill that does the job in minutes. Is it any surprise Amy was a Peace Corps volunteer?

Wired News article
Macarthur Fellows Announcement
MIT web page
Amy's course of field trips at MIT The D-Lab

Thursday, October 07, 2004

Fruitflies really do like bananas

The Nobel prize for medicine went this year to Richard Axel and Linda B. Buck who described the genes and cells of the olfactory system. 

 Interestingly, the expression of functional olfactory receptors, an important subclass of G-protein coupled receptors in tissue culture has proven difficult

As a result, the premier genetic model organism, Drosophila melanogaster, is now gaining acceptance in olfactory research. 

Recent (well, last year) research finally allows us to say with certainty that while time flies like an arrow, fruit flies like bananas.

SN-GC is the method employed: single neuron electrophysiogical recordings from living flies as detectors for gas chromatography columns.

This allowed the establishment of a preference for (more neuronal firing) odors from rotting fruit, and yes in particular from bananas. 


Human olfactory receptors have been expressed in C. elegans, so it seems likely that the functional expression of human olfactory receptors could be achieved in antennal drosophila neurons. 

If this can be shown, it would plausible that with some effort the full array of human olfactory receptors would be expressed in drosophila. 

 If not in D. melanogaster, another method will eventually be found for functional expression of hORs, and then the full complement of ligands for each receptor can be defined. 

 If flies are used, an updated descendant of the SN-GC technique could be used.

The matrix of hOR response to a particular mixture of odorants (e.g. a perfume) could then be used to define odors in support of patents and trademarks

 Confirmation that time's fun when you're having flies awaits further work. 

 Articles I couldn't shoehorn in above.


Tuesday, October 05, 2004

podcasting

Okay, I've had a chance to listen to a bit of those 720 Mb of downloaded mp3s that came in by default with installation of pyPodder. It is painful to listen to, and I couldn't find anything in there worth keeping.

Take my word, you won't be missing anything by skipping it all. I'm still interested in the concept as a way to update the compute farm, but as far as personal use is concerned, I'm not going to bother.

File distribution methods

A problem faced by all large scale compute farms is distribution and maintenance of large amounts of data. For bioinformatic compute farms, this is the sequence data that underlies all other analyses.

Do methods of distributing mp3 encoded audio such as RSS 2.0 enclosures, BitTorrent and scripted clients used in podcasting offer an alternative method for distribution?

I'm currently trying pyPodder. Works, but on my first spin it downloaded 720 Mb of spoken word mp3s. Now why is it we all need broadband again?

Other refs: RSS|ipodding

network analysis using tools from bioinformatics

Slashdot follows up on a previous article describing use of sequence analysis methods taken from bioinformatics to enable network protocol analysis.

The actual source code is now available at Protocol Informatics. They seem to be taking C++/python approach, and have a rather fresh look at bioinformatics, with a few interesting links.

I've signed up for the mailing list and downloaded the code.

Why Information Biology?

Why not bioinformatics, or computational biology, or systems biology? Information Biology is the phrase that best fits my own particular slant on computation and living systems. I was captured at an early age by the realization that DNA and proteins can be beautifully summarized as strings of alphabetic characters. This now common place representation still amazes me.

Bateson defined information best: news of difference that makes a difference. For information to exist, there must always be an entity to sense and care about differences. The world of the living is a place "... where distinctions are drawn and differences can be a cause", which is to say where information is used.

Shannon, in describing how to measure information, was careful to note that "semantic aspects of communication are irrelevant to the engineering problem", ie he avoided completely the question of meaning. In his definition the quantity of information obtainable on receipt of a single message is proportional to the number of equipropable messages that could have been sent. If only two such messages are possible, receipt of one offers the smallest possible piece of information. Note the clear differentiation of possible quantity of information from actual information. Shannon talks about how much information could be sent, not whether the message was meaningful. From the engineering perspective it doesn't matter. Unfortunately, after this careful beginning, first Shannon and then most others have used the word information to refer to this quantitative capacity for transmitting or holding meaningful information. Too bad!

Many if not most references to information are hopelessly confused due to the failure to distinguish Bateson's meaningful information from Shannon's capacity for information, resulting eventually in the sad occurrence of such oxymorons as "useless information" or "information overload" . While it's alright to call a pint of beer a pint, since no one will confuse a unit of liquid measure for a refreshing beverage, it's quite a different story with differences that make a difference as opposed to the ability to write them down in one place. No one can see the difference between a zeroed hard drive and one filled with an entire collection of music or all the secrets to the construction of advanced weapons, but the difference is real. The size of the drive determines the amount of music or number of secrets that can be stored, but of doesn't indicate whether they are there - this should be as obvious as the pint and the beer, and yet it is often and profoundly ignored in common discussion. There should really be two words, one for Bateson's information (the music and the secrets), and one for Shannon's (the size of the hard drive), but at present we have only one. Readers beware!