Functional and Evolutionary Attributes

1
Functional and Evolutionary Attributes through
Analysis of Metabolism
Sophia Tsoka European Bioinformatics Institute
Cambridge UK
2
Outline

• Metabolic reconstruction in a variety of species
• Analysis of sequence to function in metabolism
• Evolution of metabolic enzymes and pathways
• Gene fusion analysis in enzymatic sets

3
Enzyme Dataset Characterisation

• Experimental information for the entire known
  metabolic complement of E. coli
• Representation of metabolic pathways includes
• reactions
• compounds
• stoichiometry
• isozymes
• etc
• Powerful query capabilities using lisp, java

4
Metabolic Reconstruction - 1

• For genome characterisation
• Methanococcus jannaschii
• 40 of genome archaeal-specific
• 1792 proteins
• 130 pathways (22 complete)
• 297 reactions identified
• 609 reaction frames
• 461 enzymatic reaction frames
• 510 compound framesc

Tsoka, Simon, Ouzounis, Archaea, 1(4), 223, 2004
5
Metabolic Reconstruction - 2

• For drug target discovery
• Plasmodium falciparum
• 5366 proteins
• 122 pathways
• 697 reactions
• 861 enzymatic reaction
• 525 compounds
• 216 chokepoint reactions as drug targets

Yeh, Hanekamp, Tsoka, Karp, Altman, Genome
Research, 14, 917, 2004
6
Molecular and Functional Diversity of Metabolic
Pathways

• Associate enzyme sequence and function for E.
  coli small-molecule metabolism set
• Sequence enzyme families
• Function reaction types and pathways

• How many enzyme families changed in function?

• How easily is each new function invented?

• Provide insight to
• function prediction based on sequence homology
• evolution of biochemical pathways

Tsoka Ouzounis, Genome Research, 11, 1503-1510,
2001
7
Evolution of Metabolic Networks

• Taxonomic groups
• non-redundant set of all known protein sequences

• E. coli enzyme dataset
• 548 enzymes
• 208 monomers
• 348 complexes
• 132 pathways

Archaea - (21K)
Bacteria - (139K)

• Eukarya
• Protista (17K)
• Fungi (24K)
• ViridiPlantae (72K)
• Metazoa (187K)

Viruses - (122K)
Peregrin-Alvarez, Tsoka, Ouzounis, Genome
Research, 13, 422-427, 2003
8
How Conserved is Small-Molecule Metabolism across
Taxonomic Groups?
Are Other Proteins More/Less Conserved than
Metabolic Enzymes?
Bacterial enzymes are less-species specific
and more phylogenetically diverse, compared to
control samples
Peregrin-Alvarez, Tsoka, Ouzounis, Genome
Research, 13, 422-427, 2003
9
Computational Prediction of Protein Interactions

• Sequence-based prediction of protein interactions
• phylogenetic profiles (Pellegrini et al. PNAS,
  96, 4286, 1999)
• gene order (Dandekar et al. Trends Bioch. Sci.
  324, 1998)
• gene clusters (Overbeek et al. PNAS, 96, 2896,
  1999)
• gene fusion (Enright et al. Nature, 402, 86,
  1999,
• Marcotte et al.
  Science, 285, 751, 1999)

Is Fusion Frequency in E. coli Enzymes Higher
than in Other Classes of Proteins?
A Gene Fusion Event
Tsoka Ouzounis, Nature Genetics, 26, 141-142,
2000
10
Conclusions

• Analysis of sequence to function in metabolic
  networks
• Evolution of metabolic enzymes and pathways
• Genome constraints as a means to detect protein
  interactions
• Analysis of entire genomes has the potential to
  discover
• the organisational principles of cells and
• the mechanisms of their evolution

11
Acknowledgements

• Medical Research Council
• Computational Genomics Group, EBI
• Christos Ouzounis
• Jose Manuel Peregrin-Alvarez
• Leon Goldovsky
• Russ Altman and Iwei Yeh
• Peter Karp
• .. and his group