What next after the Human Genome Project

1
What next after the Human Genome Project?
Ehud Shapiro Weizmann Institute of Science
2
The talk will have three parts

• Molecular Biology in One Slide
• Why are computers so successful in some parts of
  biology?
• Why not (yet) in systems biology, and what should
  be done about it?

3
Part IMolecular Biology in One Slide
4
A computer scientists view of Molecular Biology
in one slide

• Sequence Sequence of DNA and Proteins

5
(No Transcript)
6
(No Transcript)
7
To a computer scientist Labeled 3D graphs
8
Molecular Biology in One Slide

• Sequence Sequence of DNA and Proteins
• Structure 3D Structure of Proteins and other
  biomolecules and molecular complexes

9
(No Transcript)
10
E.coli metabolism
11
S.cerevisiae protein-protein interactions
12
(No Transcript)
13
(No Transcript)
14
Molecular Biology in One Slide

• Sequence Sequence of DNA and Proteins
• Structure 3D Structure of Proteins and other
  biomolecules and biomolecular complexes
• Systems Function, activity and interaction of
  molecular systems in cells (and beyond)

15
Part IIWhy are computers so successful in
sequence and structure biology?
16
Computers are the means for consolidating
sequence biology

• Computers are key to sequence identification
• Computer data bases store accumulated sequence
  information
• Computer algorithms are used for sequence analysis

17
Computers are the means for consolidating
sequence biology

• Computers are used to is share, compare,
  criticize and correct sequence information
• The result Scientists converge to consensus
  sequences quickly and effectively

18
Computers are the means for consolidating
structural biology

• Computers are key to structure identification
• Computer data bases store accumulated structure
  information
• Computer algorithms are used for structure
  analysis

19
Computers are the means for consolidating
structural biology

• Computers are used to is share, compare,
  criticize and correct structure information
• The result Scientists converge to consensus
  structures quickly and effectively

20
Computer-based consolidation of systems biology?

• Tens of thousands of articles a year about the
  function, activity and interaction of molecular
  systems in cells
• Knowledge is encapsulated in prose, pictures and
  diagrams
• Where are the computers?

21
Computer-based consolidation of systems biology
would allow

• Handling the huge amount of accumulated knowledge
• An objective knowledge repository
• Sharing, comparing, criticizing and correcting
  accumulated knowledge
• Converging to a consensus quickly and effectively

22
Computer-based consolidation of systems biology

• The deep reason The use of good abstractions
  for sequence and structure knowledge, no
  successful abstraction (yet) for systems biology

23
What is an abstraction?

• a mapping from a real-world domain to a
  mathematical domain (homomorphism)
• highlights essential properties while ignoring
  other, complicating, ones.

24
Sequence biology uses the DNA-as-string
abstraction

• Relevant Captures sequence information, ignoring
  many biochemical properties
• Compute-able Enables string algorithms,
  efficient data-bases
• Understandable A string over A, T, C, G is the
  universal format for genetic information
• Extensible E.g., the addition of a fifth symbol
  denoting methylated cytosine.

25
Structural biology uses the Protein-as-3D-labeled
graph abstraction
26
Part IIIWhy computers are not (yet) as useful in
systems biology, and what should be done about
it?
27
Which abstraction for systems biology?
?
Which abstraction? may be the most crucial
question facing systems biology today.
28
Further reading Cellular Abstractions Cells
as Computation, Regev and Shapiro, Nature,
September 26th, 2002
29
A human engineering history perspective

• Sequence -- plowing a field
• Structure - building a house
• Systems building and flying an airplane

30
A computer science perspective

• Sequence (strings) -- 1936 50s
• Structure (graphs) - 50s 70s
• Systems (processes) 60s present and future

31
The New Biology

• The cell as an information processing device
• Cellular information processing and passing are
  carried out by networks of interacting molecules
• Ultimate understanding of the cell requires an
  information processing model
• Which?

32
Conclusions

• The most advanced tools describing computer
  systems may also be the best tools for the
  describing biomolecular systems
• Its great if we can use the decades-long effort
  in the study of concurrency in computer science
• An essential foundation for Virtual in silico
  cell project