Evaluation of Placement Techniques for DNA Probe Array Layout

1

Evaluation of Placement Techniques for DNA Probe
Array Layout
Andrew B. Kahng1 Ion I. Mandoiu2 Sherief Reda1
Xu Xu1 Alex Zelikovsky3
(1) CSE Department, University of California at
San Diego
(2) CSE Department, University of Connecticut
(3) CS Department, Georgia State University
2
Introduction to DNA microarrays and border
minimization challenges
Previous probe placement algorithm
Partitioning-based probe placement
Quantified sub-optimality of placement
Comparison of probe placement heuristics
Conclusions and future research directions
3
Introduction to DNA Probe Arrays
DNA Arrays are composed of probes where each
probe is a sequence of 25 nucleotides
Tagged fragments flushed over array
Laser activation
Images courtesy of Affymetrix.
4
A 3 X 3 array
CG
AC
G
AC
ACG
AG
AG
C
CG
Nucleotide Deposition Sequence ACG
array probes
5
A 3 X 3 array
CG
AC
G
AC
ACG
AG
AG
C
CG
C ? Mask 2
array probes
Nucleotide Deposition Sequence ACG
6
A 3 X 3 array
CG
AC
G
AC
ACG
AG
AG
C
CG
Nucleotide Deposition Sequence ACG
G ? Mask 3
array probes
A Nucleotide Deposition Sequence defines the
order of nucleotide deposition
A Probe Embedding specifies the steps it uses in
the sequence to get placed
7
Intentionally exposed sites
Border
Unwanted illumination
8
Probe Placement
? Similar probes should be placed close together
T
T
T
G
C
A
A
Deposition Sequence
T
T
G
C
C
C
A
A
T
T
Probes
T
C
C
Border 8
9
Probe Embedding
T
T
G
C
A
A
A
Deposition Sequence
T
T
T
T
G
Border 4
Border 2
C
C
C
A
T
A
Probes
A
T
C
T
T
C
Synchronous embedding deposit one nucleotide
in each group of ACGT
Asynchronous embedding no restriction
10
Probe Selection
Logic Synthesis
Probe Selection
Logic Synthesis
BIST and DFT
Analogy
Probe Placement
Placement
Physical Design
Probe Embedding
Routing
VLSI Chip
11
Give n2 probes
Find
Placement of probes in n x n sites
Embedding of the probes
Minimize
Total border cost
12
Introduction to DNA microarrays and border
minimization challenges
Previous probe placement algorithm
Partitioning-based probe placement
Quantified sub-optimality of placement
Comparison of probe placement heuristics
Conclusions and future research directions
13
Border minimization was first introduced by
Feldman and Pevzner. Gray Code masks for
sequencing by hybridization, Genomics, 1994, pp.
233-235
Work by Hannenhalli et al. gave heuristics for
the placement problem by using a TSP formulation.
Kahng et al. Border length minimization in DNA
Array Design, WABI02, suggested constructive
methods for placement and embedding
Kahng et al. Engineering a Scalable Placement
Heuristic for DNA Probe Arrays , RECOMB03,
suggested scalable placement improvement and
embedding techniques
14
Probe 1
Probe 2
Probe 3
Probe 4
Probe 1
G
C
C
C
G
C
T
T
T
C
A
T
A
A
A
G
T
G
T
G
C
T
C
C
C
A
C
A
C
A
Hamming Distance 4
Hamming Distance (P1, P2) number of
nucleotides which are different from its
counterpart border (synchronous embedding)
How to place the 1-D ordering of probes onto the
2-D chip?
15
Thread on the chip
Probe 1
2
3
G
C
4
1
A
G
C
A
Optimized Edge
Not Optimized Edge
16
(i, j)
Switch
For each site position (i, j)
Move the best probe to (i, j) and lock it in this
position
Find the best probe which minimize border
17
Introduction to DNA microarrays and border
minimization challenges
Previous probe placement algorithm
Partitioning-based probe placement
Quantified sub-optimality of placement
Comparison of probe placement heuristics
Conclusions and future research directions
18
Probe Selection
Logic Synthesis
Probe Selection
Logic Synthesis
BIST and DFT
Analogy
Partitioning
Probe Placement
Placement
Physical Design
Placement
Probe Embedding
Routing
VLSI Chip
Question Shall we use partitioning in probe
placement?
19
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
Row-Epitaxial Placement Border 48
Can partitioning based placement achieve
improvement for 25-nucleotide probes?
20
Choose a probe as seed 3 which has the largest
total Hamming distance with seed 1 and seed 2.
Choose a probe as seed 4 which has the largest
total Hamming distance with seed 1, seed 2 and
seed 3.
Choose a probe as seed 2 which has the largest
Hamming distance with seed 1.
Randomly choose a probe as seed 1.
21
Level 1 Partition
Level 2 Partition
Row epitaxial one by one
Border aware
22
Introduction to DNA microarrays and border
minimization challenges
Previous probe placement algorithm
Partitioning-based probe placement
Quantified sub-optimality of placement
Comparison of probe placement heuristics
Conclusions and future research directions
23
A
A
T
T
A
T
T
A
A
A
T
T
C
G
G
C
T
A
C
C
G
G
G
C
C
G
G
C
n4
n2
G
G
T
A
For synchronous embedding, Border 2 for any
two neighbor probes.
24
n x n real chip
Four isomorphic copies with the same border
25
Introduction to DNA microarrays and border
minimization challenges
Previous probe placement algorithm
Partitioning-based probe placement
Quantified sub-optimality of placement
Comparison of probe placement heuristics
Conclusions and future research directions
26
Chip size range between 100x100 and 500x500
Type of instances
Randomly generated
2-D Gray code
Scaled / suboptimality test cases
Embedding methods
Synchronous
Asynchronous
Quality measure
Gap from lower bound
Total border cost
CPU
Normalized cost
All tests are run on Xeon 2.4 GHz CPU.
27
Borders
Gap from lower bound
Chip size
Chip size
CPU
Normalized cost
Chip size
Chip size
Partitioning Based (Level2)
TSP Threading
Row Epitaxial
Compared with row epitaxial, new method reduce
the border cost by 3.7 and is 3 times faster.
28
Borders
TSP Threading
Row Epitaxial
Recursive Partitioning
Chip size
Gap from Optimal solution
Chip size
29
Borders
Row Epitaxial
Partitioning Based (Level2)
Chip size
Scaling ratio
Chip size
30
T
G
G
C
C
C
A
Deposition Sequence
T
T
G
C
C
A
A
Perform polishing one by one
G
C
C
Probes
C
T
A
Border 8
Border 4
Use polishing algorithm to re-embed each probe
with respect to its neighbors
31
Borders
Gap from lower bound
Chip size
Chip size
Normalized cost
CPU
Chip size
Chip size
Partitioning Based (Level2)
TSP Threading
Row Epitaxial
Compared with row epitaxial, new method reduce
the border cost by 4 and is 2.65 times faster.
32
Introduction to DNA microarrays and border
minimization challenges
Previous probe placement algorithm
Partitioning-based probe placement
Quantified sub-optimality of placement
Comparison of probe placement heuristics
Conclusions and future research directions
33
We draw a fertile analogue between DNA array and
VLSI Design Automation
We propose a new recursive partitioning-based
placement algorithm and a new embedding algorithm
which achieves 4 improvement
We study and quantify the performance of existing
and newly proposed algorithms on benchmarks with
known optimal cost as well as scaling
suboptimality experiments
34
Stronger placement operators leading to further
reduction in the border cost.
Future work also covers next generation chips 10k
10k.
Add flow-awareness to each optimization step and
introduce feedback loops.
Add the pools of probes taken from probe
selection tool.