Two Component Systems Sequence Characteristics Identification in Bacterial Genome

1
Two Component Systems Sequence Characteristics
Identification in Bacterial Genome

• Yaw-Ling Lin
• Dept Computer Sci. Info. Management,
• Providence University, Taichung, Taiwan.

2
Two-Component System

• Two-component systems (2CS)
• Sensor histidine kinase
• response regulator
• The major controlling machinery in order for
  bacteria to encounter a diverse and often hostile
  environment

3
2CS in Pseudomonas aeruginosa PAO1

• Genome 6.3M bp
• predicted genes 5570
• 123 genes were classified as 2CSs.

Complete genome sequence of Pseudomonas
aeruginosa PAO1, an opportunistic pathogen.
Nature. 2000 Aug 31406(6799)947-8. by Stover
CK, Pham XQ, Erwin AL, et al.
http//www.pseudomonas.com/
4
2CS in PAO1
5
2CS in PAO1
6
2CS in PAO1
7
2CS in PAO1

• There are 123 annotated 2CS genes in PAO1.
• Use systemic analysis of the evolutionary
  relationships between the sensor kinase and
  response regulator of a 2CS.
• Construct phylogenic trees using Clustal-W for 54
  sensor kinases and 59 response regulators.

8
2CS in PAO1 -- Sensor Tree
9
2CS Regulator Tree
10
Subtrees Analysis of 2CS
11
Co-evolution subtree Analysis
versus
SensOr Tree
Regulator Tree
12
Distance Function
13
Algorithm AGR
14
k-agreement Problem
15
k-agreement Algorithm

• Subtree sum stored the numbers of leaves on the
  internal node --- bottom-up, O(n) time
• Bucket sort the internal nodes --- O(n) time
• Identifies the subtrees of T1 with k leaves, say
  A1 to Ap. --- O(n) time
• Identifies the subtrees of T2 with k leaves, say
  B1 to Bq. --- O(n) time
• Cross checking A1 to Ap and B1 to Bq --- O(n)
  time

16
Co-evolution sensor/regulator Trees

• 46 paired clustered sensors and regulators
• Six groups of 2CS genes were identified
• Group A 2CS sensor genes PA1335, PA5166 and
  PA5511 and their linked regulator genes PA1336,
  PA5165, and PA5512 are apparently co-evolved from
  a common ancestor. The PA5165/PA5166 pair is
  happened to split earlier than the PA1335/PA1336
  and PA5511/PA5512.
• Group B, E and F each containing two 2CS operons
  are probably also evolved by gene cluster
  duplication.
• Group C contained Bordetella parapertussis
  bvgAS-like gene clusters PA3044/PA3045 and
  PA3946/PA3947/PA3948.
• Ggroup D, the gene clusters PS2571/PA2572 and
  PA2882/PA2881 appeared to be perfectly
  co-evolved.

17
Co-evolution Tree Analysis
18
Co-evolution Tree Analysis
19
Correlation analysis

• Does gene duplication tend to occur within a
  relative short distance on a bacterial genome?
• Idea a dot-matrix plot will be created, with the
  X-axis being the physical distance, and Y-axis
  being the evolutionary distance, between two
  comparing 2CS.
• Some subset of 2CS, presumably functionally
  related, could possess the correlation between
  their physical and evolutionary distances.

20
k-correlation Problem
21
k-correlation is NP-complete

• Let M1 be an adjacent matrix of a graph G, and M2
  be an zero matrix.
• If we can solve the k-correlation problem in
  polynomial time, then the maximum independent set
  problem will be polynomial solvable.

22
Conclusion

• Identifying novel 2CS in other bacteria genomes
  as well as in eucaryotic genomes
• Clustering analysis of 2CS for functional
  prediction of uncharacterized genes
• Co-evolutionary analysis of 2CS

23
Future Research

• Identifying novel 2CS in other bacteria genomes
  as well as in eucaryotic genomes
• Clustering analysis of 2CS for functional
  prediction of uncharacterized genes
• Co-evolutionary analysis of 2CS