DNA computing

1
DNA computing
2
DNA Computing

• DNA structure
• A computer science problem
• A DNA computing experiment
• Future?

3
DNA Computing

• DNA structure
• A computer science problem
• A DNA computing experiment
• Future?

4
DNA double stranded helix
5
DNA base pairs
Excerpt from Watson and Crick, Nature, 4356,
737-728 (1953)
6
DNA base pairs
7
DNA double stranded helix
Excerpts from Watson and Crick, Nature, 4356,
737-728 (1953)
8
DNA a style?
Life Science, UC Davis
Perth, Australia
Chambord, France
9
DNA Computing

• DNA structure
• A computer science problem
• A DNA computing experiment
• Future?

10
Graph Theory
The Koenigsberg bridge problem
Is it possible to find a route that crosses all
bridges only once?
11
Graph Theory
Eulers solution

• Euler noticed that, if a correct path exists
• there are two types of nodes interior, and
  Start, End
• the number of edges at an interior node must be
  even
• There can be only two nodes (Start and End) that
  have
• an odd number of edges
• The graph for Konigsberg does not satisfy these
  criteria!
• It has 4 nodes with odd number of edges

12
Graph Theory
Euler path path in a graph that goes from a
Start node to an End node by going over
all edges only once. Hamilton path path in a
graph that goes from a Start node to an End
node by going through each vertex only once.
13
DNA Computing

• DNA structure
• A computer science problem
• A DNA computing experiment
• Future?

14
Algorithm to find Hamilton paths
(Adleman, Science, 266, 1021-1024, 1994)

• Step 1 Generate random paths on the graph
• Step 2 Keep only those paths with the correct
  start and end
• Step 3 Keep only those paths with the correct
  number of vertices
• Step 4 Keep only those paths that enter each
  vertex of the graph at least once
• Step 5 If any path remains, say Yes, otherwise
  say No.

15
A specific problem
0
6
Find an Hamilton path from 0 to 6
16
A specific problem
4
3
1
0
6
2
5
Find an Hamilton path from 0 to 6
17
A Molecular Biology Solution
Encoding a graph in DNA
Vertices For each vertex i, use a random
sequence of length 20 Examples S2
TATCGGATCGGTATATCCGA S3 GCTATTCGAGCTTAAAGCTA
S4 GGCTAGGTACCAGCATGCTT Edges For an edge
i-gtj, use the sequence corresponding to the last
10 nucleotides of Si and 10 first nucleotides
of Sj. Example S2-gt3
GTATATCCGAGCTATTCGAG S3-gt4 CTTAAAGCTAGGCTAGGTA
C
18
A Molecular Biology Solution
Generating a path

• Prepare splints
• Splints are complement of the vertices.
• Example
• Vertex 3 S3 GCTATTCGAGCTTAAAGCTA
• Splint 3 cS3 CGATAAGCTCGAATTTCGAT
• Mix in solution all edges and all splints
• How two edges connect
• S2-gt3 S3-gt4
• GTATATCCGAGCTATTCGAGCTTAAAGCTAGGCTAGGTAC
• CGATAAGCTCGAATTTCGAT
• 3) Ligate DNA

cS3
19
A Molecular Biology Solution
Checking if a path is Hamilton

• Step 1 Put in solution all edges and all
  splints
• Step 2 Eliminate all paths that do not start
  with 0 and do
• not end with 6
• Amplify fragments with correct start and end.
• Step 3 Only keep paths with 140 nucleotides
• Run DNA on gel keep fragments of size 140
• Step 4 Check that DNA fragments of size 140
  contains all
• vertices
• Prepare beads, one for each splint. If fragment
• binds to bead i, it contains vertex i.
• Step 5 remaining paths are Hamilton paths

20
DNA Computing

• DNA structure
• A computer science problem
• A DNA computing experiment
• Future?

21
A computer to detect cancer
(Benenson et al, Nature, 429, 423-428, 2004)
Overview
Basic computing step
Example of application
The full procedure