Ideker, Galitski, and Hood –Systems Biology (updated 1/26/2008)

Looking at the key words. What does "discovery science" mean to you? What is the difference between a genome and a proteome?

What do you think is systems biology? Why is this new? How does this relate to more common, reductionist approaches?

What does high throughput mean?

In the "high throughput genetic manipulation" p 345, they are talking about the yeast deletion set. We have sets of yeast strains in microtiter dishes - each strain has a disruption in one gene. The disruption has a kanamycin resistance marker in it and is flanked by 2 "bar codes" that allow you to identify that mutant in a mix of mutant strains. We'll talk about this later in class.

Remember to write down your questions.

This was written in 2003 - they say that the human genome project has led us to the view that biological systems can be looked at as genes and networks of regulatory interactions. Do you think this is true? What else would you put in the equation?

Systematic gene mutations - what do you think the value of this is? What questions would you ask if you had a set of these types of mutants for your favorite organism? How does this relate to your understanding of genetics?

Why do you think this paper is so focused on tools? For me, this is an interesting review, because we've learned so much in the past 5 years! There has been a lot of tool development, more information about the fine points of eukaryotic regulation, and much more sequence information.

Computational biology - databases - these were probably the first essential computational tools (I could be lying here).

By global analysis, he means high throughput analysis. Problem then and now is that the data is sparse, even though we have 6K to 2M datapoints per assay.

Computer modeling. This is a big part of what is being done now.

Emergent properties. This is the holy grail of genomic analysis - finding something totally unexpected as a result of having so much data, especially data that gives information that can be integrated to give a multi-dimensional view of cells, tissues, etc.

Key point in talking about models on P. 353 - "The number of model parameters must be compatible with the amouint and type of available data." This is always important to keep in mind.

Framework of Systems Biology:

• Define all the components of the system
• Systematically perturb and monitor components of the system
• Reconcile the experimentally observed responses with those predicted by the model
• Design and perform new perturbation experiments to distinguish between multiple or competing model hypotheses

Interesting that the review then talks about specific regulatory networks that are known. How global is this information? Can we think of the cell as a grouping of modules each of which may be able to be modeled and the models combined?

“All of this information is hierarchical in nature:DNA -> mRNA -> protein -> protein interactions -> informational pathways -> informational networks -> cells ! tissues or networks of cells -> an organism -> populations -> ecologies. Of course, other macromolecules and small molecules also participate in these information hierarchies, but the process is driven by genes and interactions between genes and their environments.”

Forward engineering - this shows (and this can happen in a variety of ways) that you can work with the computer program or with the organism to improve models.

If you skim through the specific examples, make sure to read the summary. Have we accomplished any of this?

Other papers

Andrew Murry – Whither Genomics?

What does he mean that genetics and genomics differ in scale and focus?

Genomics gets us started thinking about modules. Does it go beyond that?

Do you think he is right about the migration of researchers to the 6 major model organisms? Go to SciSearch and see how many papers there are with yeast in the title or abstract for 1995, 2000, and 2005. Do the same for the word genome. What did you find?

Why would he think that the effect of genomics for studying eukaryotes would be different from its effect on prokaryotic biology?

What is the difference between exploratory or discovery-based science and hypothesis-driven science and why do you think that there are problems with doing both?

The real issue is how biologists deal with large datasets. Experimental design is not taught in biology – and statistics is rudimentary. Why is this a challenge for genomics?

He anticipates that proteomics is going to be hard.

How does evolution inform genomcs and genomics inform evolution?

Bader, et al. Functional Genomics to Proteomics 

Genomics has spawned a new lexicon – a million ‘omes'. There is a great deal of excitment about figuring out how life works - and finding "emergent properties"

How do you envision a “map of a cell”? Why is it hard to do this? This is "Systems Biology"

What do you need?

1. an ordered list of parts - is this hard to come by?
2. whole genome DNA sequence is the start - then you need to identify coding and non-coding regions, genome annotation
3. After whole genome is done, there is functional genomics - highthroughput, generates lots of data, integrative
   1. where are the genes?
   2. what mRNAs are abundant when (transcriptional profiling, microarrays) ?
   3. what proteins bind to what DNA sequences?
   4. what proteins are made and stable at a given time and where are they?
   5. what modifications do the proteins have (post-translational modifications) and how does that affect genome-scale functions?
   6. how do these proteins interact (genetic and protein-protein interaction)?
   7. how does a cell respond at every level to a stress (Systems Biology) ?
   8. what are the inter-level signals?
   9. can we use this information to design a cell (Synthetic Biology) ?
4. each question may require the development of completely new technology (what do you think "phenomic" analysis is?)
5. data has to be high quality and organized - databases. Go to Saccharomyces Genome Database (SGD) http://www.yeastgenome.org/ and look around. Also, check out some of the protein-interaction databases mentioned.
6. requires interactions with computer scientists, engineers, mathematicians, statisticians, and biologists
   1. you need to not be afraid to learn new things
   2. you need to learn how to talk and work with people in different disciplines
   3. you need to bring your math to as high a level as possible
   4. you need to recognize common issues that biologists have with controls, replicates, and wanting to make connections, etc.