Engineering and Understanding Biochemical Networks Herbert Sauro - PowerPoint PPT Presentation

Engineering and Understanding Biochemical Networks Herbert Sauro Bioengineering, UW Seattle

We’re trying to understand this:

Let’s start with something simpler

Let’s start with something simpler E S P

Add another step E S E S S 1 2 1 2 3 And this is where things start going downhill rapidly

Add another step E S E S S 1 2 1 2 3 Assuming S1 and S3 are fixed (at 0.5 mM and 0 mM) and steady state concentration of S2 = 0.25 mM and the Keq for each reaction is 1.5 and 0.75 respectively. Q1. Which enzyme would be the best candidate to overexpress if we wanted to maximize the flux? Q2. If we can overexpress both enzymes what is the optimal relative increase in E1 and E2 that maximizes the flux? Q3. If we can overexpress both enzymes but the total E1+E2 is fixed what is the optimal distribution in E1 and E2 that maximizes the flux? PS You’re not allowed to run a computer simulation, you may use a non-programmable calculator.

Add some more S S S S E E E 2 3 1 4 2 3 1

The reality…and it gets much worse

Solution: 1 • Understand the operating principles of kinetic networks. • Just as there is an elaborate theory of electrical circuits we need something similar for biological networks. • We teach network dynamics largely as an empirical science devoid of fundamental principles and insight. And when we do try to teach principles, sometimes what we tell our students is just wrong (check common textbooks).

Solution 1: There are good frameworks to think about metabolism (or other networks), namely Metabolic Control Analysis/Biochemical Systems Theory. And there are also advances being made in understand the economics of the cellular economy. Terrance Hwa.

Solution: 2 • Computer Models • But but… we don’t know all the parameters!

Recent Success stories These are NOT fitted models

We’re trying to understand this:

We’re trying to understand this: So why can’t we? Its not for lack of data* It’s the lack of education. Systems Biology education in Europe is more advanced that it is in the US and the emphasis tends to be different. * Although one could argue that there is a lack of the right kind of data

How will be understand/model whole cells? • Educate out students better • Use a hybrid modeling approach • Use approximations were detail doesn’t matter • Use more precision where detail does matter • Use reconstituted systems to tease out subtle but important details. • We iterate and we validate heavily. • FINALLY, let the models and theory tell use what data to collect.

Acknowledgements Andy John Doyle, Michael Savageau, David Fell, Henrick Kacser, Brian Ingalls, Mustafa Khammash, Boris Kholodenko, Frank Brugemann, Reinhart Heinrich, Jannie Hofmeyr, Jacky Snoep, Stefan Schuster, Hans Westerhoff, Athel Cornish-Bowden, Bas Tuesink Lucian Stanley