Home
Members
Best of 20062007
Best of 20072008
Past Reading
Changes

News
Next meeting
 Currently scheduled for Friday, 1/12 at 9am.
Scheduling: same time?
 Does Friday at 9am still work for everyone? If not, lets use this space to work out a new time:
Format: Discussion Points?
 Is anyone interested in trying a slightly new format? Instead of posing a question this week, each person reads the paper and responds with a discussion point for our weekly meeting. By discussion point, I mean a short, interesting thought or question, with the intention of promoting a discussion on the topic. We did this in one of Winfree's courses, and it worked really well. Post your discussion point here on or before Thursday afternoon, and we can all take a look at what has been posted thursday evening, and think them over before our meeting.
Reading
During the Engineering Principles in Bio conference at Cold Springs (which a few of us were at back in December), Uri Alon's name came up quite a bit, so it seems appropriate to do one of his papers. Here is an interesting one on the connection between fitness and regulatory mechanism, though if anyone has any other favorites, go ahead and post them here too.
 Shinar G, Dekel E, Tlusty T, and Alon U. Rules for biological regulation based on error minimization. Proc Natl Acad Sci U S A. 2006 Mar 14;103(11):39994004. DOI:10.1073/pnas.0506610103  PubMed ID:16537475  HubMed [Shinar]
 Milo R, ShenOrr S, Itzkovitz S, Kashtan N, Chklovskii D, and Alon U. Network motifs: simple building blocks of complex networks. Science. 2002 Oct 25;298(5594):8247. DOI:10.1126/science.298.5594.824  PubMed ID:12399590  HubMed [Milo]
 ShenOrr SS, Milo R, Mangan S, and Alon U. Network motifs in the transcriptional regulation network of Escherichia coli. Nat Genet. 2002 May;31(1):648. DOI:10.1038/ng881  PubMed ID:11967538  HubMed [ShenOrr]
 Dekel E and Alon U. Optimality and evolutionary tuning of the expression level of a protein. Nature. 2005 Jul 28;436(7050):58892. DOI:10.1038/nature03842  PubMed ID:16049495  HubMed [Dekel]
All Medline abstracts: PubMed  HubMed
Discussion Points
Eric
 I really like how the first paper makes a clear connection between noise minimization, fitness and regulatory mechanism. There are additional consequences of noise/error levels, and some of them are demonstrated through the modular construction of regulatory proteins (for transcriptional regulation these would have separate domains for binding the DNA and for recruiting the polymerase.)
 The paper shows how variation in the time a protein is required can cause changes in the regulatory scheme (for instance, to change a repressor to an activator). There are additional environmental factors that may change independently, for instance: the number of proteins required for a metabolic process.
 let [math]\displaystyle{ N }[/math] be the number of proteins required for [math]\displaystyle{ p }[/math] percent of the time the protein is required. [math]\displaystyle{ p }[/math] is controlled by a combination of the DNA binding energy and the number of regulatory proteins (remember the Phillips paper from week1. So while it would be possible to change [math]\displaystyle{ N }[/math] via a change in [math]\displaystyle{ p }[/math], it might have negative effect due to an increased sensitivity to errors. It would be best to change [math]\displaystyle{ N }[/math] independently of [math]\displaystyle{ p }[/math], hence the usefulness of modular domains for DNAbinding and Polymerase recuitment.
 There are certainly other uses for modularity in proteins, but I think this is interesting on a short term scale. The forth paper discusses the level of a protein is tuned fairly quickly to optimum via evolution, so I wonder if something like this might be in effect.
 what other independent environmental parameters have corresponding functional parts? can anyone think of one for the time scale over which an environment switches between conditions?
 Last year I went hear Alon talk about the development of modular logic networks under variable and modular fitness landscapes, and modular protein structures seems a direct correlary: functional separation allows for the independent tuning and erroroptimization of the different signal transfer steps of a cellular circuit.
 I've known how modularity is important to engineering and programming, but this errorminimization was an additional constraint that I hadn't considered previously. What other seemingly equivalent tradeoffs are affected by noise propogation?
